Embed Size (px)
Citation preview
POST-IMPLEMENTATION REVIEW OF THE NET METERING POLICY
IN NAMIBIA AND DESIGN OF DISTRIBUTED GENERATION HOSTING
CAPACITY ALGORITHM
Prepared by:
Markus A. Sam
SMXMAR003
Department of Electrical Engineering
University of Cape Town
Prepared for:
Dr. David Oyedokun
Department of Electrical Engineering
University of Cape Town
APRIL 2021
Submitted to the Department of Electrical Engineering at the University of Cape Town in partial
fulfilment of the academic requirements for a Master of Science degree in Electrical Engineering.
Universi
ty of
Cape T
own
The copyright of this thesis vests in the author. No quotation from it or information derived from it is to be published without full acknowledgement of the source. The thesis is to be used for private study or non-commercial research purposes only.
Published by the University of Cape Town (UCT) in terms of the non-exclusive license granted to UCT by the author.
Universi
ty of
Cape T
own
i
The copyright of this thesis vests in the author. No
quotation from it or information derived from it is to be
published without full acknowledgement of the source.
The thesis is to be used for private study or non-
commercial research purposes only.
Published by the University of Cape Town (UCT) in terms of the non-exclusive license granted to UCT by the author.
ii
DECLARATIONS
1. I, Markus A. Sam, hereby presents this dissertation in PARTIAL fulfillment of the
requirements for my Master of Science degree in Electrical Engineering.
2. I know the meaning of plagiarism and declare that all of the work in the dissertation, save
for that which is properly acknowledged, is my own.
3. This dissertation has been submitted to the Turnitin module (or equivalent similarity and
originality checking software) and I confirm that my supervisor has seen my report and any
concerns revealed by such have been resolved with my supervisor
4. I hereby grant the University of Cape Town free license to reproduce for the purpose of
research either the whole or any portion of the contents in any manner whatsoever of the
above dissertation.
Signature Date: 15/04/2021
iii
ACKNOWLEDGEMENTS
First, I would like to thank the almighty God for granting me the wisdom and the courage to
complete this work. Secondly, I wish to thank my supervisor, Dr. D.T.O. Oyedokun for his
incredible mentorship, editorial role, and selfless support to make my work a success. My
brother K. Akpeji is also acknowledged.
I further wish to thank my family; my mother, my siblings, and most importantly, my special
person R. Amunyela for the unwavering support in my academic endeavours. My friends: P.
Niipale, I. Matheus, M. Nepando, and L. Abed are also acknowledged.
Finally, I would like to express my deepest gratitude to the management of Erongo RED, under
the leadership of Mr. Fessor Mbango, for permitting me to use Erongo RED’s data in this work.
The support and assistance rendered by Erongo RED’s staff (Mr. G. Thiel, Mr. F. Nandiinotya,
Mr. M. Jankowski, Ms. H. Ndamaala, and Mr. E. Engombe) is acknowledged.
iv
ABSTRACT
Global campaigns against fossil fuels to reduce the emission of greenhouse gases and combat
climate change has compelled the electricity supply industry (ESI) around the globe to explore
environmentally friendly sources of electricity. The concept of distributed generation (DG) has
gained momentum and is emerging as a promising source of clean energy, with immense
potential to maximize the shares of renewable energy in the global energy mix. Like the rest of
the world, Namibia is witnessing an unprecedented growth of DG courtesy of governmental
efforts to ensure a speedy transition to low-carbon generation technologies. In 2016, the
Namibia government developed the net metering (NM) policy known as the Net Metering Rules
(NMR) as a consumer-focused approach to achieve the low-carbon objective. To date, there
has been no rigorous post-implementation review of the NMR to assess its effectiveness,
despite rising concerns from distribution network operators (DNOs) about whether the NMR is
suited for long-term application in the fast-growing market of prosumers.
This study conducts a broad appraisal of the status quo on DG integration into distribution
networks in Namibia and an in-depth assessment of the technical and financial impacts of the
NMR using the Erongo Regional Electricity Distributor (Erongo RED) as a case study. The
findings indicate that most prosumers export over 60% of generated energy to distribution
networks and achieve significant financial savings by offsetting on-site demand with their
generation in real-time, as well as, by offsetting a portion of their electricity bills through NM
compensations for grid exports. NM compensations at the avoided cost makes grid exports in
Namibia a cheaper alternative source of energy to DNOs as compared to the national utility,
which charges other energy service charges i.e. reliability charge, transmission losses charge
etc. on top of the avoided cost. Additional findings indicate that prosumers are subjecting DNOs
to revenue losses because of reduced volumetric energy sales caused by the reduction of
prosumers’ on-site energy requirements from the grid. With the deployment of DG growing
rapidly in Namibia, increasing grid exports and associated technical constraints are envisaged
in distribution networks.
This dissertation recommends adaptations to existing regulatory policies to mitigate envisaged
financial and technical risks associated with DGs. These adaptions include a DG hosting
capacity (HC) assessment methodology for consumer-side photovoltaic (PV) DG in existing
v
distribution network where a high and uniform uptake of DG is anticipated. The methodology
captures time dependency correlations between load and generation profiles, which increase
the accuracy of HC results. The uniqueness of the methodology is the concept of calculating
monthly HC, which aids the optimal integration of DG into distribution networks to meet
consumers’ daily energy requirements throughout the year without comprising the network’s
quality of supply. The methodology was tested on a residential and business distribution
network. Results confirm that HC in distribution networks varies monthly. However, the practical
implementation of monthly HC require upgrades to existing inverter technology, which currently
contains a single export limit functionality. This opens the possibility to drive innovation in the
inverter technology, to develop a date-based multiple export limit functionality. The results also
demonstrated the importance of considering phase unbalance when conducting HC studies for
residential distribution networks. Applications and limitations of the methodology were
discussed.
vi
TABLE OF CONTENTS
DECLARATIONS .................................................................................................................................... ii
ACKNOWLEDGEMENTS .................................................................................................................... iii
ABSTRACT ............................................................................................................................................ iv
LIST OF ABBREVIATIONS ................................................................................................................. ix
LIST OF FIGURES ................................................................................................................................. x
LIST OF TABLES ................................................................................................................................. xii
1. INTRODUCTION............................................................................................................................ 1
1.1 Background Information ....................................................................................................... 1
1.2 Problem Statements .............................................................................................................. 3
1.3 Research Hypothesis ............................................................................................................ 3
1.4 Research Questions .............................................................................................................. 3
1.5 Research Objectives ............................................................................................................. 4
1.6 Scope and Limitations........................................................................................................... 4
1.7 Research Paper under Review ........................................................................................... 4
1.8 Dissertation Outline ............................................................................................................... 4
2. LITERATURE REVIEW ................................................................................................................ 7
2.1 Electricity Supply Industry (ESI) in Namibia ...................................................................... 7
2.1.1 Structure of the ESI in Namibia ................................................................................... 7
2.1.2 Electricity Supply in Namibia ....................................................................................... 8
2.1.3 Electricity Demand in Namibia ..................................................................................... 9
2.1.4 Government Interventions to Unlock Potentials for Renewable Energy ............. 10
2.1.5 Developments on Utility-Side and Consumer-Side Distributed Generations (DG)
11
2.2 Evolution of Distribution Networks .................................................................................... 13
2.2.1 Passive Distribution Networks ................................................................................... 13
2.2.2 Active Distribution Networks ...................................................................................... 15
2.3 Definition of DG .................................................................................................................... 15
2.3.1 Distributed Energy Resources (DER) and Technologies ...................................... 16
2.3.2 Benefits of DG .............................................................................................................. 24
vii
2.3.3 Challenges of DG ........................................................................................................ 25
2.3.4 Importance of Power Flow Studies in Networks with DG ...................................... 29
2.3.5 DG Integration Standards and Practices ................................................................. 30
2.4 Hosting Capacity (HC) of DG in Distribution Networks .................................................. 33
2.4.1 HC Definition ................................................................................................................ 33
2.4.2 Benefits of Assessing HC ........................................................................................... 34
2.4.3 HC Methodologies and Metrics ................................................................................. 34
2.4.4 HC Practices in Different Countries .......................................................................... 37
2.5 DG Compensation Mechanisms and Incentive Policies ................................................ 39
2.6 Energy Storage Technologies and Applications ............................................................. 41
2.6.1 Types of Energy Storage Technologies ................................................................... 41
2.6.2 Applications of Energy Storage Systems ................................................................. 43
2.7 Role of Inverter Export Limit Function to Meet Grid Code Requirements .................. 44
2.8 DigSilent Programming Language (DPL) ........................................................................ 45
3. POST-IMPLEMENTATION REVIEW OF THE NAMIBIAN NET METERING RULES
(NMR) .................................................................................................................................................... 46
3.1 Provisions of the Net Metering Rules (NMR) .................................................................. 46
3.2 The motivation of the NMR Review .................................................................................. 47
3.3 Choice of Case Study for the NMR Review .................................................................... 47
3.4 Focus and Considerations of the NMR Review .............................................................. 50
3.5 Results and discussions ..................................................................................................... 52
3.5.1 Technical analysis results .......................................................................................... 52
3.5.2 Financial analysis results ........................................................................................... 55
3.6 Highlights of the review....................................................................................................... 57
3.7 Policy imperatives ................................................................................................................ 58
4. DEVELOPMENT OF METHODOLOGY FOR ASSESSING HOSTING CAPACITIES (HC)
IN EXISTING LV DISTRIBUTION NETWORKS ............................................................................. 60
4.1 The motivation of HC Methodology .................................................................................. 60
4.2 Considerations of HC Methodology .................................................................................. 61
4.2.1 Modelling Approach ..................................................................................................... 62
4.2.2 Distribution Network Model ........................................................................................ 63
4.2.3 Consumer Load Model ............................................................................................... 63
4.2.4 PV System Model ........................................................................................................ 64
viii
4.2.5 Performance Indices ................................................................................................... 65
4.3 Proposed Methodology ....................................................................................................... 66
5. TESTING AND RESULTS OF METHODOLOGY FOR ASSESSING HOSTING
CAPACITIES ........................................................................................................................................ 69
5.1 Summary of the Global Horizontal Irradiance (GHI) of Walvis Bay ............................. 69
5.2 Testing HC Methodology on !Nara Residential Distribution Network (Test 1) ........... 70
5.2.1 As-Built Drawing of !Nara Network ........................................................................... 70
5.2.2 DigSilent Model of !Nara Network in DigSilent PowerFactory 2018 .................... 72
5.2.3 Simulation Results and Discussion of the HC results of !Nara Network (Test 1)
74
5.3 Testing HC Methodology on !Nara Residential Distribution Network (Test 2) ........... 77
5.3.1 Results (Test 2) ............................................................................................................ 77
5.3.2 Discussions (Test 2) .................................................................................................... 79
5.4 Testing HC methodology on a business distribution network (Test 3) ........................ 80
5.4.1 As-Built Drawing of Spar Network ............................................................................. 80
5.4.2 Model of Spar Network in DigSilent PowerFactory 2018 ...................................... 82
5.4.3 Simulation Results and Discussion of the DG HC of Spar Network .................... 84
5.5 Applications of HC Results ................................................................................................. 86
5.6 The relevance of Developed HC Methodology ............................................................... 87
5.7 Limitation of HC Methodology and Results ..................................................................... 87
6. CONCLUSION.............................................................................................................................. 89
7. RECOMMENDATIONS ............................................................................................................... 94
REFERENCES .................................................................................................................................... 97
APPENDICES .................................................................................................................................... 111
APPENDIX A .................................................................................................................................. 111
APPENDIX B .................................................................................................................................. 114
APPENDIX C .................................................................................................................................. 117
APPENDIX D .................................................................................................................................. 118
APPENDIX E .................................................................................................................................. 119
APPENDIX F .................................................................................................................................. 120
ix
LIST OF ABBREVIATIONS
ADMD After Diversity Maximum Demand
CHP Combined heat and power
DER Distributed Energy Resources
DG Distributed Generation
DNO Distribution Network Operator
DPL DigSilent Programming Language
ECB Electricity Control Board
ESI Electricity Supply Industry
FiT Feed-in Tariff
GSF Generation Scaling Factor
HC Hosting Capacity
IPP Independent Power Producer
MD Maximum Demand
MME Ministry of Mines and Energy
NDP5 Fifth National Development Plan
NEP National Energy Policy
NIRP National Integrated Resource Plan
NM Net Metering
NMD Notified Maximum Demand
NMR Net Metering Rules
PPA Power Purchase Agreement
PV Photovoltaic
RED Regional Electricity Distributor
SADC Southern African Development Community
SANS South African National Standard
SAPP Southern Africa Power Pool
TOU Time of Use
WECS Wind energy conversion systems
ZESCO Zambia Electricity Supply Corporation
ZPC Zimbabwe Power Corporation
x
LIST OF FIGURES
Fig. 1: Structure of Namibia’s electricity supply industry [27], [28]. .......................................... 8
Fig. 2: Statistics for Namibia’s energy imports from SAPP .................................................... 10
Fig. 3: Distributed energy resources (DER) technologies ...................................................... 17
Fig. 4: A simplified energy band diagram at T > 0 K for a direct bandgap (EG) semiconductor
[66] ....................................................................................................................................... 18
Fig. 5: Single-diode model of the theoretical PV cell and equivalent circuit of a practical PV
device [65], [66]. ................................................................................................................... 19
Fig. 6: PV hierarchy. ............................................................................................................. 20
Fig. 7: Global renewable generation capacity [68] ................................................................. 21
Fig. 8: WECS general structure [69], [70] .............................................................................. 21
Fig. 9: Biomass Conversion Technologies [73], [74] ............................................................. 23
Fig. 10: Two-bus distribution system without DG .................................................................. 26
Fig. 11: Two-bus distribution system with DG ....................................................................... 27
Fig. 12: Number of prosumer DG connected to Erongo RED’s grid per year ......................... 49
Fig. 13: Capacity of prosumer DG connected to Erongo RED’s grid per year ....................... 49
Fig. 14: A prosumer’s daily import/export profile ................................................................... 50
Fig. 15: Monthly distribution of the occurrence of prosumers’ annual peak export power ...... 54
Fig. 16: Hourly distribution of the occurrence of prosumers’ annual peak export power ........ 54
Fig. 17: Daily distribution of the occurrence of prosumers’ annual peak export power........... 55
Fig. 18: Residential prosumers (energy import value vs energy export value) ....................... 56
Fig. 19: Business Prosumers (energy import value vs energy export value) ......................... 56
Fig. 20: General modeling approach ..................................................................................... 63
Fig. 21: Modeling and hosting capacity assessment algorithm. ............................................. 68
Fig. 22: Monthly global horizontal irradiance (GHI) for Walvis Bay ........................................ 69
Fig. 23: Cadastral layout of !Nara network ............................................................................ 70
Fig. 24: !Nara distribution network configuration and properties ............................................ 71
Fig. 25: DigSilent network model for !Nara network .............................................................. 73
Fig. 26: Inside configuration of a distribution kiosk model ..................................................... 74
Fig. 27: Script summary report for !Nara network during June 2019 (Test 1) ........................ 74
Fig. 28: Network and Consumer level monthly HC results for !Nara network (Test 1) ........... 75
Fig. 29: Script summary report for !Nara network during June 2019 (Test 2) ........................ 78
xi
Fig. 30: Network and Consumer level monthly HC results for !Nara network (Test 2) ........... 78
Fig. 31: Cadastral layout of Spar network ............................................................................. 81
Fig. 32: Network configuration of Spar network ..................................................................... 81
Fig. 33: DigSilent Network Model for Spar Network .............................................................. 83
Fig. 34: Inside configuration of a meter-room model ............................................................. 84
xii
LIST OF TABLES
Table 1: Power Stations in Namibia [31] ................................................................................. 9
Table 2: Namibia’s energy trading agreements [19], [28] ........................................................ 9
Table 3: Utility-side DG developments [30] ........................................................................... 12
Table 4: Consumer-side DG developments .......................................................................... 13
Table 5: Response to abnormal voltages (IEEE) [79] ............................................................ 30
Table 6: Response to abnormal voltages (IEC) [78] .............................................................. 31
Table 7: Current distortion limits [78], [79] ............................................................................. 31
Table 8: Business (Three-Phase) tariffs [133] ....................................................................... 51
Table 9: Performance indices and limits................................................................................ 66
Table 10: Distribution of single-phase loads on phases ........................................................ 71
Table 11: Technical constraints encountered on !Nara network (Test 1) ............................... 75
Table 12: Technical constraints encountered on !Nara network (Test 2) ............................... 78
Table 13: Properties of LV feeder cables in Spar Network .................................................... 82
Table 14: NMD and PV capacity for consumers in Spar Network .......................................... 83
Table 15: Hosting capacity results of Spar network ............................................................... 84
1
1. INTRODUCTION
1.1 Background Information
The mounting global pressure against the use of fossil fuels to reduce the emission of
greenhouse gases, combat climate change, and reduce environmental impacts has compelled
the electricity supply industry (ESI) around the world to explore environmentally friendly
sources of generation. Distributed generation (DG) is emerging as a promising alternative of
centralized generation. Its adoption is playing a crucial role in ensuring the transition from fossil
fuels to low-carbon generation technologies, ultimately maximizing the take of renewable
energy resources in the global energy mix [1]–[7].
Countries across the globe have implemented subsidized programs to promote and accelerate
the adoption of DG through various incentive approaches. These incentives are generally
provided via feed-in tariffs (FiT) and net metering (NM) policies [1], [7]–[14]. Incentivized
programs have played a significant role in the proliferation of DG in distribution networks [4].
Responding to the growing demand for consumer-side DG in the country, the Namibian
government, in fulfilling its obligation to combat climate change, developed a NM policy, also
known as the Net Metering Rules (NMR), to incentivize the adoption of consumer-side DG.
This policy was enacted in 2016 to provide a uniform regulatory platform for consumer-side
DG. Before the promulgation of the NM policy, distribution network operators (DNOs) in
Namibia had little practical experience on DG integration because it was still an emerging
concept. Yet, since its implementation there has been no rigorous post-implementation review
of its effectiveness, despite rising concerns from DNOs in Namibia and several other countries
about whether NM policies are suited for long-term application in the fast-growing market of
prosumers [15]. Yael and Benjamin [16] defined prosumers as electricity consumers that both
consume and produce electricity.
The growing number of DG installations and overwhelming demand to access distribution
networks for DG deployment requires DNOs to have a good understanding of hosting capacity
(HC) concepts. HC entails the maximum amount of DG that could be integrated into a
distribution network without violating normal operational limits [17]. A distribution network’s HC
for DG depends on numerous factors including, inter alia, load profiles, generation profiles, the
2
concentration of DG, network configuration, voltage levels, network characteristics (shared or
dedicated), and climatic conditions, especially for weather-dependent DG such as solar
photovoltaic (PV) and wind [2], [4], [6], [11], [18]. These factors should be considered inclusively
when conducting technical studies aimed at assessing hosting capacities of DGs in distribution
networks. In Namibia, solar PV, a weather-dependent distributed energy resource (DER), is
the predominant DER employed for consumer-side DG due to the abundant solar irradiation
levels in the country, believed to be the second-highest in the world [19]. When conducting HC
assessment studies, it is vital to consider methodologies that can manage time dependency
correlations between load and generation profiles, to capture the uncertainties caused by the
intermittency of PV power generation [20].
This dissertation conducts a broad appraisal of DG integration in the networks of Namibian
DNOs, and an in-depth assessment of the technical and financial impacts of the NMR using
Erongo Regional Electricity Distributor (Erongo RED) as a case study. Based on the qualitative
and quantitative findings from the study, adaptations to current regulatory policies on DG
integration in Namibia are recommended to encourage a fair, inclusive, and sustainable
deployment of consumer-side DGs, with a reduced risk of violating normal operating conditions
of distribution networks. As part of the adaptions, this dissertation develops and tests a script
based HC assessment methodology for consumer-side DGs in existing distribution networks
where a high and uniform uptake of DG is anticipated. This is a data-driven methodology that
requires real distribution network parameters, historical time-series load profiles, and solar
irradiance data for estimating PV power generation profiles. The methodology calculates the
HC of a distribution network at both consumer and network levels. Uniquely, the methodology
introduces a concept of calculating HC on a monthly basis, given metering data for specific
months. Monthly HC presents DNOs the opportunity to manage grid exports on a monthly basis
by imposing curtailment requirements based on a network’s monthly adequacy to host DGs.
This concept aims to acknowledge the influence of monthly weather variations on load
consumption and solar irradiation, which have a direct impact on HC. This methodology also
acknowledges the presence of existing DG systems on distribution networks.
Other Sub-Saharan African (SSA) countries might find the results, discussions, and
recommendations in this study very helpful in informing their own DG integration policies.
3
1.2 Problem Statements
Namibian DNOs had little practical experience on DG integration before the promulgation of
the NMR because it was still an emerging concept. Since its implementation, there has been
no rigorous post-implementation review to assess its effectiveness, despite rising concerns
among Namibian DNOs about whether the NMR is suited for long-term application.
Several studies have been conducted to assess the impacts of DG on the Namibian national
grid, particularly impacts posed on network stability and losses [21]–[23]. Namibia, however,
lacks published studies focused on the impacts of DG on distribution networks. To fill this
research gap, studies need to be conducted to foster an understanding of the benefits and
impacts of DG, as well as, hosting capacities (HC) of DG in distribution networks, in the
Namibian context.
1.3 Research Hypothesis
The understanding of HC in distribution networks will enable DNOs to develop innovative and
cost-effective approaches to manage and allocate network capacity to consumers fairly and
inclusively, such that all consumers are considered as prospective prosumers, especially in
areas where a high and uniform uptake of DG is expected. Deployments of DGs within the HC
of distribution networks significantly alleviate the risks of technical challenges in distribution
networks. The research hypothesis, therefore, states that:
“A DG capacity dissemination methodology that considers consumer inclusivity and grid
reliability provide a platform for all electricity consumers in Namibia to participate in DG
initiatives, such as net metering, in a sustainable and non-discriminatory manner.”
1.4 Research Questions
The following research questions are the fundamental drivers of this study and the yardsticks
upon which its success is measured:
a) What is the status quo of net metering in Namibia?
b) What are the impacts of consumer-side DG in distribution networks?
c) What adaptations to current utility and regulatory policies are required to encourage a
sustainable deployment of consumer-side DG?
4
d) What methodology can DNOs use to assess the DG hosting capacity of their distribution
networks?
e) How can DNOs utilize DG hosting capacity results to implement operational strategies to
ensure a sustainable integration of consumer-side DG?
1.5 Research Objectives
This study seeks to achieve the following objectives:
a) Conduct a post-implementation review of the NMR to assess the performance and impacts
of consumer-side DG.
b) Develop and test a methodology to assess the DG hosting capacity in Namibian distribution
networks. This methodology shall be aligned to methodologies developed and used in other
parts of the world.
c) Propose adaptations to the current NM policy and recommend effective operation strategies
to establish a sustainable DG regulatory platform.
1.6 Scope and Limitations
The focus of this study is narrowed to consumer-side DG within the regulatory scope of the
Namibian NMR. Utility-side DG connected to distribution networks by independent power
producers (IPP) through power purchase agreements (PPA) are excluded from the scope of
this study.
1.7 Research Paper under Review
[1] M. A.Sam, D.T.O Oyedokun, and K.O. Akpeji, "Design of Distributed Generation Hosting
Capacity Algorithm”, Journal of Energy of Southern Africa, 2020.
Status: The paper is still under review
1.8 Dissertation Outline
This report is structured as follows:
5
Chapter 1 introduces the dissertation by discussing the background information, problem
statement, hypothesis, objectives, scope, and limitations of the study.
Chapter 2 concisely reviews the literature on the structures of the electricity supply industry
(ESI), the evolution of distribution networks, distributed generation (DG), DG hosting capacity
(HC) of distribution networks, DG compensation mechanisms, and incentive policies, energy
storage technologies and applications, inverter export limit function to meet grid code
requirements, and DigSilent Programming Language (DPL) to set a broad context for this study.
Chapter 3 reviews the provisions of the Namibian net metering rules (NMR) and discusses the
motivations, choice of case study, focus, and considerations of the post-implementation review
of the Namibian NMR. It also presents and discusses the technical and financial results of the
post-implementation review. Highlights and policy imperatives from the review are also
discussed in this chapter.
Chapter 4 discusses the motivations, considerations (in terms of the modelling approach,
network model, load model, PV system model, and performance indices), and development of
the script-based methodology for assessing LV distribution networks’ DG hosting capacity.
Chapter 5 discusses the testing of the methodology developed in chapter 4 on two real LV
distribution networks (residential and business). It also presents and discusses the results of
each test. Applications of results, limitations of the methodology, and relevance of the
methodology are also discussed in this chapter.
Chapter 6 summarizes the findings of the study and provides answers to the research
questions.
Chapter 7 recommends necessary adaptions required by DNOs and the regulator to embrace
consumer-side DG without compromising the reliability of distribution networks or putting the
revenues of DNOs at stake.
References, a list of references cited in chapters 1 to 7 are presented here.
6
Appendices, which contain extra information supplementing results already presented in
chapter 5 are included at the end of the dissertation.
7
2. LITERATURE REVIEW
2.1 Electricity Supply Industry (ESI) in Namibia
2.1.1 Structure of the ESI in Namibia
The electricity supply industry (ESI) in Namibia plays a pivotal role in the development and
economic growth of the country because most economic sectors, such as manufacturing,
mining, enterprises, agriculture, fisheries, etc require electricity for their operations [24].
Namibia’s Fifth National Development Plan (NDP5) described electricity as the “indispensable
force that drives all economic activities in Namibia” [24]. This certainly means that the socio-
economic development of the country relies on the reliability, accessibility, and affordability of
the electricity supplied by the ESI. The Namibian ESI is comprised of three main components,
namely, generation, transmission, and distribution. The ESI also features energy trading –
imports and exports of energy from and to the Southern African Development Community
(SADC) through the Southern Africa Power Pool (SAPP) [25], [26].
Fig. 1 encapsulates the structure of the Electricity Supply Industry (ESI) in Namibia. The
Ministry of Mines and Energy (MME) is the leading body of the energy sector in Namibia,
followed by the Electricity Control Board (ECB), which is the regulating authority of the
Electricity Supply Industry (ESI). The national utility, Namibia Power Cooperation (NamPower)
comes third in the hierarchy after the ECB. NamPower is the custodian of generation,
transmission, and energy trading. The generation component also shelters the independent
power producers (IPPs) as generators of renewable energy. The fourth level after NamPower
encompasses the role players in the distribution industry, distribution network operators
(DNOs) namely, the Regional Electricity Distributors (RED) and local authorities (municipalities
and town councils) [27], [28].
8
Fig. 1: Structure of Namibia’s electricity supply industry [27], [28].
2.1.2 Electricity Supply in Namibia
Namibia has a total installed generation capacity of 626.8 MW, which includes diesel, coal, and
hydro generation plants, as well as, other operational renewable energy-based plants, mainly
wind and solar photovoltaic (PV) [29], [30]. According to Table 1, Ruacana Power Station, with
an installed capacity of 347 MW, is the current biggest power plant in the country. The second-
largest installed capacity of 137.3 MW comes from utility-side distributed generations (DG)
mainly employing solar PV and wind plants. The remaining installed capacity comes from Van
Eck Power Station and Anixas Station, which are operated as standby power stations during
peak hours [31].
Namibia is an operating member of the SAPP. This pool allows member countries to import
and export energy from and to other members. Table 2 shows four Namibian active energy
trading agreements with three SAPP members. Two agreements are with South Africa through
9
Eskom, one is with Zambia through Zambia Electricity Supply Corporation (ZESCO), and the
last is with Zimbabwe through Zimbabwe Power Corporation (ZPC).
Table 1: Power Stations in Namibia [31]
Name Type Installed Capacity Operating Mode
Ruacana power
station
Run-of-the-river
hydro 347 MW Flexible
Van Eck power
station Coal 120 MW Emergency Stand-by
Anixas power station Diesel/heavy fuel oil 22.5 MW Emergency Stand-By
Renewable energy
plants
(solar PV and wind)
Solar and Wind 137.3 MW Flexible
Table 2: Namibia’s energy trading agreements [19], [28]
Import Source Max Supply Capacity
(MW)
Capacity Factor
(%)
Net Supply
(MW)
South Africa
(Eskom supplemental) 200 20% 40
South Africa
(Eskom off-peak) 300 50% 150
Zimbabwe (ZPC) 80 50% 40
Zambia (ZESCO) 50 100% 50
TOTAL 630 MW 280 MW
2.1.3 Electricity Demand in Namibia
Namibia is characterized by a lack of baseload generation. To date, the local supply in the
country relies on the Ruacana Power Station; its primary local source of generation. The
performance of this plant depends on the availability of water flow in the Kunene River [31].
Lack of water in the river due to poor rainfalls often affect the performance of this plant and its
contribution to the local energy mix [31]; this inevitably causes an increase in energy imports
10
from neighbouring countries [19], [27]. In 2017, critically low water flow was observed in the
Kunene River due to a poor rainfall observed in the southern part of Angola. This negatively
affected the performance of the plant and increased the country’s energy imports from
neighbouring countries through SAPP, consequently setting a new record of 73% of energy
imports. A new maximum demand of 653 MW was also recorded in the same year, which was
higher than the maximum demand of 631 MW recorded in the previous year [31]. According to
NamPower [25], [26], [31]–[38], Namibia had imported on average, 61% of the country’s annual
electricity demand from neighboring countries in the past 10 years, as shown in Fig. 2. The
statistics evidently show that Namibia’s local generation still cannot meet the local demand,
hence justifying the reliance on foreign energy imports to fill the deficit and ensure the security
of supply.
Fig. 2: Statistics for Namibia’s energy imports from SAPP
2.1.4 Government Interventions to Unlock Potentials for Renewable Energy
Namibia has great potential for renewable energy, which could help establish the most needed
local baseload generation in the country or perhaps reduce reliance on electricity imports.
Firstly, Namibia is endowed with the world’s second-highest solar irradiation of up to 7
kWh/m2/day of global horizontal irradiance and up to 8 kWh/m2/day of direct normal irradiance
[19]. Secondly, the country boasts great potential for wind energy along its coast, with wind
speed reaching 10 m/s and greater [19]. Thirdly, the country has a moderate potential for
biomass from encroacher bushes [19]. Namibia also has a hydropower station located in
Ruacana, on the Kunene River, which is the primary local source of power in the country [19].
56% 54% 53%
61% 59% 58%
68%
57%
73% 71%
61%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Ener
gy Im
po
rts
fro
m S
AP
P(%
)
11
To unlock potentials for renewable energy, the Namibian government developed four policies
outlining key strategic frameworks and guidelines to diversify the country’s energy mix and
strengthen local generation capacity. These policies include the National Energy Policy (NEP)
[27], National Renewable Energy Policy (NREP) [19], National Integrated Resource Plan
(NIRP) [28], and the NM policy, also known as the Net Metering Rules (NMR) [39]. The general
theme of these policies commonly emphasize the importance of rededicating the country’s
effort towards achieving the following objectives [19], [27], [28], [39]:
a) Enhance the country’s security of supply through the utilization of local resources.
b) Reduce reliance on electricity imports.
c) Promote the generation of electricity using environmentally friendly energy resources
such as renewable energy.
d) Maximizing the take of renewable energy in the country’s energy mix.
e) Ensure local participation in generation projects.
f) Improve interconnections with neighboring countries to enhance energy-trading
opportunities.
g) Expand and increase access to affordable electricity.
h) Create income-generating opportunities locally.
2.1.5 Developments on Utility-Side and Consumer-Side Distributed Generations (DG)
Namibia had witnessed substantial developments of both utility-side DG and consumer-side
DG in the past ten years, mostly employing solar (PV) and wind energy technologies. To date,
Namibia has committed about twenty utility-side DG with a combined installed capacity of 137.3
MW (Table 3) [30]. A utility-side DG in the context of this study is defined as a DG connected
directly to a utility’s network, at voltage level of 11 kV and above, through a power purchase
agreement (PPA) with an independent power producer (IPP). Namibia have also seen an
unprecedented growth of consumer-side DG. According to the information obtained from DNOs
in Namibia, the current developments on consumer-side DG stands at 881 installations with a
combined installed capacity of 40.94 MW, as shown in Table 4.
A consumer-side DG in the context of this study is defined as a DG connected behind the meter
of an electricity consumer for self-generation. Consumer-side DG was spurred by the net
12
metering rules (NMR), feed-in tariffs (FiT) (mostly in Erongo Region), and increasing consumer
preferences for self-generation resulting from high electricity tariffs. In summary, Namibia has
a total installed capacity of 178.24 MW of DG (both utility and consumer-side DG). These
developments have enhanced the take of renewable energy in the energy mix of the country.
Utilities in Namibia include the generation & transmission national utility (NamPower) and
DNOs, which include the authorities (municipalities and town councils) and regional electricity
distributors (RED) such as Erongo RED, NoRED, and CenoRED. Their contributions to DG
developments are highlighted in Table 3 and Table 4.
Table 3: Utility-side DG developments [30]
No
Commercial licensed
renewable energy
plant
Installed
capacity
(MW)
DER
technology Off-taker utility Status
1 Osona Sun Energy (Pty)
Ltd 5 MW PV NamPower Operational
2 Omburu Sun Energy (Pty)
Ltd 4.5 MW PV NamPower Operational
3 OLC Arandis (Pty) Ltd 3.8 MW PV Erongo RED Operational
4 Metdecci Energy
Investment (Pty) Ltd 5 MW PV NamPower Operational
5 Momentous Solar One
(Pty) Ltd 5 MW PV NamPower Operational
6 HOPSOL1 Power
Generation (Pty) Ltd 5 MW PV CenoRED Operational
7 HOPSOL2 Power
Generation (Pty) Ltd 5 MW PV NamPower Operational
8 Camelthorn Business
Venture No Two (Pty) Ltd 5 MW PV NamPower Operational
9 GreeNam1 Electricity (Pty)
Ltd 10 MW PV NamPower Operational
10 GreeNam2 Electricity (Pty)
Ltd 10 MW PV NamPower Operational
11 Sertum Energy (Pty) Ltd 5 MW PV NamPower Operational
12 Alten Solar Power
(Hardap) (Pty) Ltd 37 MW PV NamPower Operational
13 Aloe Investment No. 27
(Pty) Ltd 5 MW PV NamPower Operational
14 Ombepo Energy (Pty) Ltd 5 MW Wind NamPower Operational
13
15 Ejuva One Solar Energy
(Pty) Ltd 5 MW PV NamPower Operational
16 Ejuva Two Solar Energy
(Pty) Ltd 5 MW PV NamPower Operational
17 Sun EQ Four Investment
(Pty) Ltd 5 MW PV NamPower Operational
18 B2Gold (Pty) Ltd 7 MW PV
Off-Grid B2Gold (Pty) Ltd Operational
19 ALCON Consulting
Services (Pty) Ltd 5 MW PV NamPower Operational
20 Wind Generator 0.22 MW Wind Erongo RED Operational
TOTAL CAPACITY (MW) 137.3 MW
Table 4: Consumer-side DG developments
Utility name Number of
consumer-side DG
Capacity of consumer-side
DG (MW)
DER
technology
Erongo RED 211 6.65 MW PV
CenoRED 128 4.65 MW PV
NoRED 47 4.08 MW PV
Municipality of
Keetmanshoop 37 1.56 MW PV
City of
Windhoek 458
24 MW
PV
TOTAL 881 40.94 MW
2.2 Evolution of Distribution Networks
2.2.1 Passive Distribution Networks
Traditionally, distribution networks were designed and operated as stable passive systems,
facilitating a unidirectional flow of current from centralized generation plants to loads [40]–[42].
The design of distribution networks essentially consisted of load estimation, sizing, and
selection of electrical equipment to construct a network that supplies sufficient and safe
electricity to intended consumers [43], [44]. Other factors equally considered when designing
14
distribution networks include the choice of voltage level, classification of consumers, calculation
of voltage drops, choice of network configuration (e.g. radial, ring, or meshed network), choice
of protection devices, and adapting of design to geographical and climatic conditions of the
network location [43]–[46]. The aforementioned considerations ensure that the design is safe,
stable, reliable, efficient, and cost-effective [44], [47].
Load estimation has been historically conducted using pre-determined empirical metrics that
acknowledge the existence of diversity in electricity consumption [44], [47]. These metrics
include, inter alia, diversity factor, coincidence factor, utilization factor, load factor, demand
factor, maximum demand (MD), and after diversity maximum demand (ADMD) [43], [44], [47].
The prevailing and commonly used metric is ADMD [47]. ADMD is used in distribution network
design to estimate the peak coincident load for a defined number of consumers [47]. Peak
coincident load is used to size equipment such as transformers, cables, and switch gears.
ADMD is expressed as the ratio of the simultaneous maximum demand of a group of
homogeneous consumers over the number of consumers supplied by the network [43], [45].
The design of distribution networks needs to comply with standards that define the minimum
and maximum limits within which they are required to operate. DNOs in Namibia adopted two
South African standards namely, the NRS 034-1 [44] and NRS 048 [48] for the planning and
design of distribution networks. The development of the NRS 034-1 standard adopted a
comprehensive method for calculating voltage drops in low-voltage residential feeders called
the Herman-Beta method [44]. Limits contemplated by these standards were set with the
consideration of diversity in the structures of the distribution networks arising from varying load
densities, population dispersion, and network topologies. ADMD ranging from 0.5 - 4.5 kVA,
depending on consumer classification, has been recommended for residential distribution
network designs [44]. The standards defined 400 V as the standard voltage for consumers
supplied with three-phase LV connections, and 230 V for single-phase LV connections [44],
[48]. Distribution networks are required to operate within a voltage variation of ± 10% of the
nominal voltage [44], [48]. Parallel to the two aforementioned standards, Namibia has also
adopted series of South African National Standards (SANS), which mainly specify the
performance requirements of specific electrical equipment, such as transformers, conductors,
cables, switchgear, etc.
15
2.2.2 Active Distribution Networks
Climatic and environmental concerns, as well as, campaigns against the usage of fossil fuels
gave rise to the need to harness renewable energy resources [1]–[7], [49]–[51]. To fulfill this
mandate, governments around the globe developed policies to facilitate the diversification and
incorporation of low-carbon energy resources in their energy mix [49]–[52]. Distributed
generation (DG) emerged as a promising approach to achieve the low-carbon footprint goal in
the electricity supply industry (ESI) [41], [49]–[53], and as a result, the ESI is witnessing an
unprecedented growth of DG as transition efforts from fossil fuels to low-carbon generation
technologies advances [1]–[7].
Even though distribution networks were initially designed to facilitate a unidirectional power
flow from centralized generation plants to loads [40]–[42], the emergence of DG has changed
the status quo; generation now occurs closer to the loads. This has evolved distribution
networks from passive to active networks, which are characterized by bidirectional power flow
[40]–[42]. In 2000, Ackermann et al. [51] predicted that DG would become an important part of
the future generation systems; indeed, the prediction has become a reality today. DG has been
praised for providing significant benefits to both consumers and DNOs [15], [40]–[42], [54], [55].
However, it has also been criticized for the challenges it potentially poses on the operation of
distribution networks and revenues of DNOs [53], [55], [56].
2.3 Definition of DG
Terms such as embedded generation, dispersed generation, and decentralized generation are
often used interchangeably with DG. There is yet no universal definition of DG; factors such as
location, capacity, voltage level, and technology of generation system have played a vital role
in the formulation of DG definitions by various researchers [49], [57], [58].
Ackermann et al. [51] defined DG as “an electric power source connected directly to the
distribution network or on the customer side of the meter”. Zhao et al. [59] defined DG as
“relatively small generation systems that are designed, installed, and operated in distribution
networks or distributed at the customer side to meet special customer needs and support the
operation of distribution networks based on economic, efficient, convenient, and reliable
generation”. Leaning on the definition by Ackermann et al. [51], Yamujala et al. [60] defined DG
as “the integrated use of small generation units directly connected to a distribution network or
16
on the consumer side of the meter”. Ioan et al. [61] defined DG as “the production of power
near or at the consumption place using cogeneration units and renewable energy sources”.
The focus of this study is on renewable energy-based DG, primarily photovoltaic (PV) based
DG on the customer side of the meter. DG connected directly to distribution networks are also
emphasized in this work, even though limitedly. This research adopts the DG definitions by
Ackermann et al. [51] and Ioan et al. [61] and defines DG as a source of electric power utilizing
renewable energy resources as fuel and connected directly to a distribution network or on the
customer side of the meter.
2.3.1 Distributed Energy Resources (DER) and Technologies
Chowdhury et al. [42] defined DER as electricity generation sources that are employed in the
concept of DG. These resources can be classified as either renewable energy resources or
non-renewable resources (conventional sources that use fossil fuels), as well as dispatchable
resources or non-dispatchable resources [4], [42], [57], [60], [62].
A wide range of technologies is employed to harness energy from DER. The applications of
these technologies vary with the nature of DER. The available DER technologies include
reciprocating engines, gas turbines, and microturbines, combined heat and power (CHP)
systems, wind energy conversion systems (WECS), solar (PV) systems, small-scale
hydroelectric generation, fuel cells and storage devices [42], [57], [59]. Fig. 3 summarizes a
wide range of DER technologies employed for DG [42], [54], [57], [59], [60], [62], [63].
17
Fig. 3: Distributed energy resources (DER) technologies
DER technologies applicable and viable for Namibia are mainly solar, wind, and biomass
technologies, owing to the rich solar irradiation, promising wind potential, and abundant
encroacher bush for biomass in the country [28]. For this reason, only technologies related to
these three DER are discussed in detail.
2.3.1.1 Solar Photovoltaic (PV) System
Solar PV is the predominant DER technology in Namibia and other parts of Africa that have the
rich solar irradiation [64]. This technology employs the basic principle of the photoelectric effect
to convert sunlight into electricity. The technology uses PV cells to generate direct current (DC)
DER
Technologies
Fossil fuel technologies Non-Fossil fuel technologies
Micro turbine (Gas turbine)
Reciprocating engines (Internal combustion engine)
(Diesel and gas engines)
Combustion engines
(Gas turbines)
Electrochemical energy resources (Fuel cells)
Polymer electrolyte membrane
fuel cell (PEMFC)
Alkaline fuel cell (AFC)
Direct methanol fuel cell (DMFC)
Phosphoric acid fuel cell (PAFC)
Molten carbonate fuel cell (MCFC)
Solid oxide fuel cell (SOFC)
Solar photovoltaic (PV)
Wind energy
conversion systems
(WECS)
Small-scale hydroelectric
generation
Storage energy resources
Batteries
Flywheels
Supercapacitors
Biomass conversion
technologies
Solar thermal
18
from PV cells when illuminated with light. A PV cell is therefore the elementary device of a PV
system. PV cells consist of a semiconductor diode containing the p-n junctions [65], [66].
The semiconductors used to make PV cells contain electrons with weak bonds occupying an
energy band called the valence band, as indicated in Fig. 4. If energy beyond a threshold called
bandgap energy is applied to a PV cell, the electrons in the valence band gain energy, start
vibrating, break their weak bonds and start moving freely in a new energy band called the
conduction band [65], [66]. The free electrons in the conduction band are separated from the
valence band by the bandgap (measured in units of electron volts or eV) [65], [66]. The energy
needed to free the electron can be supplied by photons, which are particles of light.
Fig. 4: A simplified energy band diagram at T > 0 K for a direct bandgap (EG) semiconductor
[66]
Where:
- EG = Direct band gap
- EC = Conduction band
- EV = Valence band
- E = Electron energies
- p = Crystal momentum
19
Fig. 5 shows a general equivalent circuit of a PV cell. This circuit is composed of a photo-
current source, diode, parallel resistor expressing the leakage current, and series resistor
describing the internal resistance to the current flow [65], [66].
Fig. 5: Single-diode model of the theoretical PV cell and equivalent circuit of a practical PV
device [65], [66].
Using Kirchoff’s current law, the I-V characteristic of the PV cell based on the equivalent cell
shown Fig. 5 is expressed mathematically in equation (1).
I= Ipv-Io [e(V+RsI
αVt)-1] -
V+RsI
Rp
(1)
Where:
- I = Net cell current
- Ipv = Photovoltaic current generated by the incident light
- Id = Diode current
- IO = the reverse saturation or leakage current of the diode
- Vt = thermal voltage
- V = Output voltage
- RS = Series resistance
- Rp= Parallel resistance
- α = Diode ideality constant
20
A group of PV cells connected in series forms solar modules, whereas a group of solar modules
connected in series and parallel form solar arrays [65], [66]. Fig. 6 adopted from [66], illustrates
the PV hierarchy showing the difference between a PV cell, PV module, and PV array.
Fig. 6: PV hierarchy.
Technologies used to manufacture PV modules are divided into two categories, namely,
crystalline modules and thin-film modules as described below.
(a) Crystalline modules
Crystalline modules are further divided into two categories namely, mono-crystalline modules
and poly-crystalline modules [42], [67]. Mono-crystalline modules have efficiency in the range
of 16 – 18%, which is perceived as the best performance for solar modules, hence this type of
module is regarded as the best module but is expensive [42], [67]. Poly-crystalline modules are
the commonly used modules, with efficiency between 12% –14%. They are typically cheaper
than mono-crystalline modules [42], [67].
(b) Thin-film modules
These are modules with a thin homogenous layer rather than a crystal structure. They have the
lowest efficiency, i.e. 6 – 8%. They are also the cheapest available modules. They are further
divided into 3 categories: amorphous silicon (a-Si) thin film, CdTe (cadmium telluride), and CIS
(copper indium selenide) [42], [67].
2.3.1.2 Wind energy conversion systems (WECS)
The renewable energy statistics presented by the International Renewable Energy Agency
(IRENA) in March 2019 show that the global renewable generation capacity has reached 2,351
21
GW. Wind energy accounts for 24% of this global renewable generation capacity, making it the
second-highest contribution after hydro energy [68]. The chart in Fig. 7 below shows the
percentage take of hydro, wind, solar, and other sources (e.g. bioenergy and geothermal) in
the global renewable generation capacity.
The wind energy conversion system (WECS) refers to the system that converts wind energy
into electrical energy [42]. The principal component in the WECS is the wind turbine, which
captures the kinetic energy of wind flow through rotor blades and converts it into mechanical
energy [42], [69]. The mechanical energy is then transferred to the generator side through the
gearbox, which drives and rotates the generator shaft to produce electricity [42], [69]. Fig. 8
shows the general structure of the WECS. The structure depicts the interconnection of various
WECS components, namely the aerodynamic, mechanical, and electrical components. The
electrical aspect of WECS can further be divided into three main components, which are wind
turbine generators (WTGs), power electronic converters (PECs), and the utility grid [70].
Fig. 7: Global renewable generation capacity [68]
Fig. 8: WECS general structure [69], [70]
22
According to Chowdhury et al. [1], the generation power output of a wind turbine depends on
several factors such as wind velocity, size, and shape of the turbine. The power developed is
expressed mathematically using (2) below [1]:
P= 1
2CpρV
3A (2)
Where:
- P = Power (W)
- Cp = power coefficient
- ρ = Air density (kg/m3)
- V = wind velocity (m/s)
- A = swept area of rotor blades (m2).
2.3.1.3 Biomass Conversion Technologies
Biomass energy is the energy produced or generated using organic materials such as bushes,
trees, crops, animal wastes, etc. [71]. To date, biomass is the only renewable energy resource
that can provide energy in all states, solid, liquid, and gas [71], [72]. It is also the only renewable
resource that is sufficiently available to be considered as a substitute for fossil fuel [71], [72].
There are three available technologies for converting biomass into energy, namely the
thermochemical conversion route, biological conversion route, and physical conversion route
[71]–[73]. The thermochemical conversion route comprises of three techniques, namely the
direct combustion method, pyrolysis, and gasification. The biological conversion route involves
fermentation and hydrolysis, whereas the physical conversion route includes densifications and
extractions [71], [73]. Thermo-chemical conversion techniques are the most prominent
techniques for converting biomass to heat and electricity. They employ processes similar to
those used with fossil fuels [73]. The three thermochemical conversion techniques are
encapsulated below and depicted in Fig. 9:
23
a) Direct combustion method: This technique involves burning biomass to produce heat
and electricity, which is commonly referred to as, a combined heat and power (CHP)
system. This includes producing steam to rotate the turbine and subsequently drive the
generator that produces electricity [71]–[74].
b) Pyrolysis conversion technique: This technique involves heating biomass in the
absence of oxygen to produce liquid products (bio-oil), gaseous products, and solid
products (charcoal).
c) Gasification conversion technique: This technique involves converting biomass in an
atmosphere of steam and air to produce a gaseous fuel, mainly by exposing solid fuels
to high temperatures and limited oxygen. The process includes reacting biomass with
air, oxygen, or steam to produce Syngas (a gaseous mixture consisting of CO, CO2,
H2, CH4, and N2) [71]–[74].
Fig. 9: Biomass Conversion Technologies [73], [74]
24
2.3.2 Benefits of DG
The ESI around the globe is witnessing an unprecedented growth of DG as efforts to shift away
from fossil fuels to low-carbon technologies advance [1]–[7]. Apart from a low-carbon footprint,
DG can provide significant benefits in the areas of power losses, voltage regulation, power
quality, and economics. Most benefits are often realized at well-regulated penetration levels of
DG [11], [40].
2.3.2.1 Utility-aligned benefits
a) Eco-friendly and clean energy
The exploitation of distributed energy resources (DER) allows DNOs to generate eco-
friendly and clean energy than that from centralized fossil-fuel generation plants; this
reduces environmental pollution and global warming effects [42], [59].
b) Transmission and distribution (T&D) losses and congestion reduction:
Power losses in a power system are proportional to the system’s impedance and magnitude
of current following through it. Reducing any of this parameter (impedance or current) can
subsequently reduce losses in a network. As opposed to centralized generations, DG
systems are in proximity to the loads, thus the energy generated by these systems is
supplied directly to the loads [3], [75]. This reduces energy requirements from the grid and
the current drawn from T&D networks, thus reducing T&D losses. Reduced energy
requirement from the grid also reduces the loading of equipment (feeders and
transformers), abating T&D congestion. Reduced losses significantly increases the overall
efficiency of T&D networks [3], [4], [11], [53], [60]–[62], [75].
c) Defer investments to upgrade existing generation and T&D infrastructures:
The ability of DG to reduce energy requirements from centralized generations and T&D
networks eliminates the need to upgrade or build new generation and T&D infrastructures
[4], [61], [62].
d) Improve feeder voltage profiles:
The integration of DG reduces the loading of distribution networks thus offsetting the
voltage drops initially caused by heavy loads. This ultimately improve the overall voltage
profiles of feeders [3], [60]–[62], [75], [76].
25
2.3.2.2 Consumer-aligned benefits
Consumer-side DG allows consumers to become energy independent [9]. DG also provide
significant energy cost savings and power supply reliability especially in areas with chronic grid
electricity interruptions. Cost savings have been achieved through self-energy reliance (less
consumption from the grid reduces electricity bill) and utility compensations via incentives, such
as net metering (NM) and feed-in tariffs (FiT), from grid energy exports [15]. Consumer-side
DG also enables consumers to reduce their demand charges through “peak load shaving”,
especially during instances where peak demand coincides with peak generation [42], [53], [54].
2.3.3 Challenges of DG
Despite the overwhelming benefits of DG, the increasing adoption of DG presents unique
economic and technical challenges to DNOs [15], [40]–[42], [53], [55]. Economic challenges
include loss of gross utility revenue due to reduced energy sales, the additional cost to conduct
technical studies, and the capital costs to upgrade infrastructures overloaded by DG [15], [55].
Technical challenges associated with DG are more prominent in the areas of voltage regulation,
thermal ratings of equipment, power quality, and protection. The degree to which DG affects
these network parameters depends on numerous factors, which include, inter alia, size and
location of DG, the concentration of DG, network load profile, connection type, length and sizes
of conductors, network configuration, voltage level, characteristics of the network (shared or
dedicated), and climatic conditions especially for weather-dependent DG [2], [4], [6], [11], [18].
Briefly, the impacts of DG are network and location-specific. Some of the challenges of DG are
briefly discussed below:
a) Impact on current flow, thermal limits, and losses
The integration of DG into distribution networks changes distribution networks from passive
to active networks, which are characterized by bidirectional power flow [40]–[42]. High
penetration of DG can result in high reverse power flow into the grid, which causes overload
problems for electrical infrastructures (cable, transformers, etc.) and increases T&D losses.
b) Impact on voltage regulation
26
One of the benefits associated with DG is its ability to support and improve voltage profiles
[62]. This benefit normally manifests at well-regulated or optimum DG penetration levels
[11], [40]. In a network characterized by low demand, this benefit could be outweighed by
high penetration levels of DG. High DG penetration causes surplus power from DG hosts
to flow back to the grid, creating reverse power flow problems and ultimately causing
voltage rise issues [9], [18], [75]–[77]. Depending on the magnitude of this reverse power
flow, the voltage may rise beyond statutory limits [18], [62], [76], which may affect the utility’s
obligation to supply power within the stipulations of quality of supply standards.
The impact of DG on voltage rise is illustrated mathematically using the voltage-drop
concept equations derived from a two-bus circuit shown in Fig. 10 and Fig. 11. This
illustration was adapted from [62], [76].
i. Two-bus distribution system without DG
Fig. 10: Two-bus distribution system without DG
Fig. 10 shows a two-bus network, V1 is the sending-end voltage while V2 is the receiving-
end voltage. PL and Q L are real and reactive components of the connected load,
respectively. Using Kirchhoff's Voltage Law (KVL), the voltage in the circuit is distributed as
expressed in (3). The overall power supplied by the circuit is expressed in (4). The current
flowing in the circuit is expressed in (5). Voltage-drop in the circuit is expressed in (6), which
is simplified in stages as expressed in (7), (8), and (9).
V1 = V2
+ I (R+jX) (3)
P+jQ= V1 * I *
(4)
I =P − jQ
V1
(5)
27
∆V = V1 − V2
= ( P−jQ
V1
) (R+jX)
(6)
∆V = V1 – V2
=PR+QX
V1
+j ( PX− QR
V1
) (7)
When the angle between the sending-end voltage and receiving-end voltage is very small,
the voltage drop in the circuit is approximately equal to the real part of the complex voltage-
drop. Additionally, if the sending-end bus were considered as the reference bus, the angle
of the sending-end voltage 𝑉1 will be zero (0), hence 𝑉1
= |𝑉1| = 𝑉1 . The voltage-drop
equation is hence simplified to (8).
∆V≈ V1 − V2≈PR+QX
V1 (8)
If equation (8) is converted to the per-unit system, with 𝑉1 considered as the base voltage
then (8) can be rewritten as expressed in (9).
∆V ≈ PR+QX (9)
ii. Two-bus distribution system with DG
To analyze the impact of DG on voltage rise, a DG is introduced to the two-bus network as
shown in Fig. 11.
Fig. 11: Two-bus distribution system with DG
When a DG is exporting power to the network, the direction of power flow is reversed and
voltage at the point of connection (POC) will increase above the sending-end voltage. The
voltage at the PoC is expected to be higher than other nodes in the network. With DG
connected to the network, (8) can be rewritten as expressed in (10).
28
∆V ≈ V1 − V2≈ R(PG-PL)
V1+
X(QG-QL)
V1 (10)
If both the load and DG are operating at the power factors of one, then (QG
-QL) will be zero,
which eliminates the contribution of reactive energy to voltage change. Equation (10) can
therefore be written as expressed in (11). Most DG operates at unity power factor as
recommended by standards for DG integration to power systems [18], [78], [79].
∆V ≈ V1 − V2≈R(PG − PL)
V1
(11)
Equation (11) mathematically shows that voltage in a network with DG systems is directly
proportional to the amount of power exported to the grid by the DG systems [62], [76]. The
impact of DG on voltage regulation is therefore justified.
c) Impact on power quality
Other literature works have described the impact of DG on power quality in terms of the
following power system disturbances:
i. Overvoltage / voltage rise
High DG penetration causes surplus power from DG hosts to flow back to the grid,
creating reverse power flow problems and ultimately causing voltage rise issues
[18], [76]. Depending on the magnitude of this reverse power flow, the voltage may
rise beyond statutory limits [53], [62], [76].
ii. Voltage flickers / fluctuations
The inherent intermittency of DER causes rapid changes in DG outputs, which
eventually causes noticeable voltage changes and subsequently voltage flickers [5],
[9], [60], [76], [80].
iii. Phase voltage imbalance
The increase of single-phase DG in distribution networks causes voltage imbalance
problems among phases [5], [9], [60], [76].
iv. Harmonic distortions
29
DG technologies using power electronics converters, such as the inverters used in
PV DG, produces voltage and current distortions, which if injected in distribution
networks affects the life span of electrical equipment and increase grid losses [2]
[5], [9], [60], [62], [75]–[77], [81].
d) Impact on protection
Increasing DG in distribution networks poses the risk of increased fault currents, blinding of
protection devices due to loss of sensitivity, false tripping, loss of selectivity, and loss of
coordination among protection devices [2], [11], [50], [52], [53], [64], [75], [77].
e) Impacts on utility gross revenue
The growing adoption of consumer-side DG has become a major threat to utility business
models [9], [82]. Consumers with DG behind the meter are becoming energy independent
therefore consuming less from the grid [9]. Through incentives such as net energy metering
and net billing, consumers with DG benefits from compensations towards their grid exports,
which DNOs use to offset their grid consumption costs [15]. Despite the benefits to
consumers, these transactions affect and subject DNOs to major gross revenue losses [15],
[55].
2.3.4 Importance of Power Flow Studies in Networks with DG
Widen [58] emphasized the importance of power flow studies in power systems. Power flow
studies provide information regarding bus voltages and power flowing through power system
lines, transformers, and other elements of a power system for a specified load demand [83].
With the grid-integration of DG, power flow studies make it possible to analyse how distribution
networks respond to bidirectional power flow caused by DG with regard to power losses,
voltage variation, protection, line flow, etc. [58].
30
2.3.5 DG Integration Standards and Practices
2.3.5.1 International Standards
To address the anticipated technical challenges associated with DG integration, several
standards, and guidelines have been developed to provide minimum requirements for
interconnecting DG systems. One of the international standards for DG integration is IEEE
1547 [79]. The second international standard is IEC 61727 [78], which was developed
specifically to guide the integration of PV systems below 10 kVA. Other international standards
include IEC 61000 series and IEEE 519. Requirements from the IEEE 1547 [79] and IEC 61727
[78] were reviewed and summarized in this report. These requirements cover mainly power
quality aspects, namely voltage, flicker, frequency, harmonics, and power factor, as
summarised below:
a) Voltage Regulation: The standards have permitted DG systems to supply current to
distribution networks within predefined voltage variation limits, as presented in Table 5
and Table 6. However, the standards have denied DG systems the right to regulate
network voltage [78], [79].
Table 5: Response to abnormal voltages (IEEE) [79]
Voltage Range
(% of Nominal Voltage) Maximum Trip Time / Clearing Time (s)
V < 50 % 0.16 s
50 % ≤ V < 88 % 2.0 s
88 % ≤ V ≤ 110 % Continuous operation
110 % < V < 120 % 1.0 s
V ≥ 120 % 0.16 s
31
Table 6: Response to abnormal voltages (IEC) [78]
Voltage Range
(% of Nominal Voltage) Maximum Trip Time / Clearing Time (s)
V < 50 % 0.1 s
50 % ≤ V < 85 % 2.0 s
85 % ≤ V ≤ 110 % Continuous operation
110 % < V < 135 % 2.0 s
135 % ≤ V 0.05 s
b) Power factor: The standards require DG systems (especially PV system) to have a
lagging power factor greater than 0.9 when the output is greater than 50 % of the rated
DG output power. Hence, most PV inverters (designed for utility-interconnected
systems) operate close to the unity power factor [78].
c) Harmonics and waveform distortion: The standards require DG systems to exhibit low
current-distortion levels to minimize the risk of affecting other equipment connected to
the distribution network. Total harmonic current distortions have been limited to less
than 5 % of the DG rated output. The individual harmonic percentage limits are provided
in Table 7. Even harmonics are limited to 25% of the odd harmonic limits [78], [79].
Table 7: Current distortion limits [78], [79]
Odd harmonics Distortion limit
3rd through 9th Less than 4.0 %
11th through 15th Less than 2.0 %
17th through 21st Less than 1.5 %
23rd through 33rd Less than 0.6 %
Even harmonics Distortion limit
2nd through 8th Less than 1.0 %
10th through 32nd Less than 0.5 %
32
d) Flicker: The operation of DG systems should not cause voltage flickers to increase
above the limits specified by the IEC 61000-3 series [78], [79].
e) Loss of utility voltage: To prevent islanding, the standards require DG systems to de-
energize the network upon loss or collapse of network voltage [78], [79].
f) Over/under frequency: A deviation from the rated frequency (50 Hz or 60 Hz) should be
kept within a range of ± 2%. If the frequency goes outside this range, the standards
require DG systems to stop energizing the network within 0.5s [78], [79].
g) Islanding protection: The standards further require DG systems to stop energizing the
network within 2s upon loss of utility network [78], [79].
h) DC injection: Under any operating condition, the injection of DC current into the
distribution network from static power converters should be limited to 1% of the static
power converter’s rated output [78], [79].
2.3.5.2 Integration practices for consumer-side DG in different countries
Integration practices for consumer-side DG in a few selected countries were reviewed to
establish the analogy of the requirements, and how they compare to requirements in Namibia.
The countries reviewed include South Africa, the United Kingdom, Australia, and the United
States of America (USA).
Integration requirements for consumer-side DG systems in South Africa are provided in the
standard documents called the NRS 097-2-1 [84] and NRS 097-2-3 [26], which highlights the
utility interface requirements and connecting criteria, respectively. The Energy Networks
Association (ENA) in the United Kingdom (UK) provides integration requirements in the
standard documents called Engineering Recommendation G98, Engineering Recommendation
G99, and Engineering Recommendation G100 [86]–[88]. The Energy Networks Australia (ENA)
in Australia provides integration requirements in a standard document called the Technical
Guidelines for Embedded Generation (EG) Connections [89], [90]. In America, the Public Utility
Commission of Texas provides integration requirements in a standard document called a DG
Distributed Interconnection Manual [91]. In Namibia, the regulator, Electricity Control Board
33
(ECB), provides integration requirements for consumer-side DG in a standard document called
the Net Metering Rules (NMR) [39].
These aforementioned standard documents reference key integration requirements provided
by the international standards [78], [79]. They also incorporate generation capacity
requirements and limits, which vary from country to country. Generation capacity limits referred
to as hosting capacities (HC) are discussed in 2.4.
2.4 Hosting Capacity (HC) of DG in Distribution Networks
Increasing demand for access to distribution networks for DG deployment requires DNOs to
explore innovatively cost-effective solutions to manage grid capacities and maximize DG
penetration without deteriorating normal operating conditions of distribution networks [20], [92],
[93]. To achieve this objective, DNOs need to study the impacts of DG on their networks and
subsequently determine possible measures for mitigating these challenges to ensure a safe
and reliable operation of distribution networks [94]. Some of the mitigation measures for DG
impacts include network reinforcements and upgrade [17], [95], network reconfiguration [1],
[96], energy storage [1], [96], and installation of on-load tap changers (OLTC) [1], [96]. These
measures require additional capital investment, hence they are less cost-effective [1], [17], [92],
[96]. Alternative solutions for mitigating DG impacts based on active power curtailment and grid
export limits were proposed in [1] and [96], respectively. Another alternative solution to mitigate
DG impacts includes limiting DG penetration to the hosting capacities of distribution networks
[17], [93]. All identified alternative solutions are effective and require low capital costs.
2.4.1 HC Definition
There is no universal definition for hosting capacity (HC), however, narratives from literature
allude that it is the maximum amount of DG that maintains the operation of a distribution
network in a satisfactory operating condition [93], [94], [96], [97]. HC has been defined as the
maximum amount of generation that can be accommodated by the distribution network without
causing adverse impacts [94]. Alternatively, HC can be defined as the maximum capacity of
DG that can be connected to a distribution network without violating its operational constraints
[95].
34
A distribution network’s HC depends on numerous factors including, inter alia, load profiles,
generation profiles, the concentration of DG, network configuration, voltage levels, network
characteristics (shared or dedicated), and climatic conditions, especially for weather-dependent
DG such as solar photovoltaic (PV) and wind [2], [4], [6], [11], [18]. If properly understood, the
concept of HC has the potential to help DNOs make informed decisions when processing DG
interconnection requests, and subsequently reduce the risk of disrupting normal operating
conditions of distribution networks, mainly voltage and thermal limit constraints [93].
2.4.2 Benefits of Assessing HC
Determining the HC of a distribution network can help DNOs achieve the following objectives:
a) Identify maximum allowable DG integration required to make informed decisions when
processing DG interconnection requests [93], [98], [99].
b) Evaluating existing networks on their adequacy to host DGs to enable DNOs and
regulators in drawing up HC guidelines [98], [100].
c) Develop new principle for designing and planning new distribution networks [100],
[101].
d) Allow DNOs to make transparent and fair decisions towards prosumers [98], [99].
e) Enhance utilization of electrical infrastructures to avoid major and costly network
upgrades [98].
f) Investigate potential adverse impacts of increasing DG connection beyond HC [98].
g) Maintain safe and reliable operation of distribution networks by reducing the risk of
disruptions to normal operating conditions [93], [98], [99].
2.4.3 HC Methodologies and Metrics
Various methodologies have been employed to evaluate the impacts of DGs in distribution
networks and subsequently determine their HC. Literatures have expressed that popularity of
two methods, namely deterministic and probabilistic method, as one of the key methods
employed in assessing HC of distribution networks [100], [102], [103]. These two methods have
been widely compared to assess their strengths and weaknesses, and determine their
application suitability. The choice of a methodology depends on the availability of network and
35
load data, and certainty of generation capacity, and location/placement of DG systems [20],
[100].
2.4.3.1 HC Methodologies
Chihota and Bekker [100] compared the deterministic and probabilistic methods in the context
of DG modelling. Traditional deterministic methods employ simulations based on singular, fixed
values, which is applied uniformly without consideration of uncertainties in consumer loads,
and variability of renewable energy based DGs. They also lacks the ability to analyze the
uncertainty in capacity and location of DGs. A probabilistic method, on the other hand, have
the ability to predict the stochasticity of loads and variability in generation, and uncertainties
pertaining the sizes and location of DGs. Voltage magnitude, phase voltage unbalance and
thermal limits of equipment are considered as possible performance metrics.
Fang et al. [102] described a deterministic method as an approach in which a number of
selected scenarios with minimum and maximum values and DG output are considered as the
non-coincidental and coincidental inputs of the analysis. A probabilistic method employs time-
series of simultaneous hourly variations of all input parameters, providing their probability
distribution.
Mulenga et al. [20] compared the strengths and weaknesses of three methods used for HC
studies of solar PV in low-voltage distribution networks, namely the deterministic method,
probabilistic/stochastic method, and time-series method. Deterministic methods apply known
fixed loading and generation input data to analyze the impacts of DG without considering
uncertainties. Stochastic methods apply unknown loading and generation input data to analyze
DG impacts with the consideration of uncertainties in both loading and generation. Time-series
methods consider certainty in loading and uncertainty in the generation, especially for PV
generation. Time-series methods utilize actual historical load measurements (known or certain
data) and PV generation calculated according to historical solar irradiance data (uncertain
data). These data often cover at least one year, divided into equal intervals (15 minutes, 30
minutes, or an hour) to capture seasonal variations in both loading and generation. Time-series
methods include time correlations between loading and generation, hence they are better
suited for time-varying assessment of hosting capacity [20].
36
In [94], a location-specific HC methodology is proposed. This methodology employs an iterative
search algorithm that browses through possible locations to examine DG impacts. The HC at
each location is obtained through a search algorithm that evaluates the parameters of the
network and ensures that performance indices are not violated, as the DG system size is
incremented with a set increment. Voltage and loading of equipment (lines and transformers)
were used as performance indices.
In [93], a scenario-based time-series simulation methodology was used to assess the impacts
and HC of a distribution network with PV based DG. This methodology requires a real
distribution network modeled with time-series historical load profiles and generation profiles
estimated using irradiance data. The PV system size is proportional to the consumer’s
maximum load. The methodology employs a particle swarm optimization (PSO) algorithm to
minimize total losses and voltage variation. Voltage and losses were used as performance
indices.
In [104], an HC methodology that uses a forward/backward sweep power flow solution was
presented. The methodology employs a search algorithm to determine all possible sizes and
locations to examine the effects of PV DG. The methodology uses peak and off-peak load
information. Voltage and loading of equipment (lines and transformers) were used as
performance indices.
In [105], an HC methodology that considers realistic time-series load, and a PV profile for a
summer day was presented. The methodology uses a search algorithm to evaluate the network
parameters against the performance indices. Voltage and loading of equipment (lines and
transformers) were used as performance indices.
In [97], an HC methodology that considers hourly time-series load and generation data
spanning over a full year was proposed. The methodology uses a search algorithm to evaluate
the network parameters against the performance indices. Voltage and loading of equipment
were used as performance indices.
The reviewed literature presents non-uniform interpretations of the probabilistic method. Some
describes it in the context of the uncertainty in load and generation data only, while others
describes it in the context of uncertainties of capacity and location of DGs, or combined
37
uncertainty of load, generation, capacity and location of DGs. Nevertheless, similarities in the
descriptions of deterministic methods by different literatures were observed. The time-series
method was identified as an separate method for assessing DG impacts and HC [20], and as
a component of the probabilistic method [102]. Nonetheless, the time-series method based on
historical network data and predicated generation have the ability to eliminate the worst case,
fixed values, and single scenario consideration of deterministic method, and also address the
uncertainty of load/generation profiles.
2.4.3.2 Performance indices and metrics
Generally, network parameter conditions are considered as performance indices or metrics for
HC assessments. The choice of an index depends on the use case of the HC results and the
intended goal of implementation. Network parameter conditions include voltage magnitude
(undervoltage and overvoltage), voltage unbalances, rapid voltage changes, voltage, and
current harmonics, interharmonics, supraharmonics, thermal overload, protection and losses
[20], [93], [94], [97], [99], [104], [105]. To ensure that distribution networks continue to operate
within design limits after integration of DG systems, performance indices for hosting capacities
are normally set based on design limits informed by network planning and design standards,
and DG interconnection standards. Voltage magnitude and loading of equipment (lines and
transformers) are the most commonly used indices for HC assessments [20], [93], [94], [97],
[104], [105].
2.4.4 HC Practices in Different Countries
A summary of HC methodologies and requirements (focused at both consumer and network
levels) for some countries is detailed below:
a) In South Africa, generation capacity requirements are defined according to the type of
network a consumer is connected. The generation capacity limit for a consumer connected
in a shared network is 25% of the notified maximum demand (NMD), while the generation
capacity limit for a consumer connected on a dedicated network is 75% of the NMD. The
NMD is often determined by the consumer’s service circuit-breaker rating [85]. A consumer
with a single-phase service connection of 60A (13.8 kW) is eligible for a DG with a capacity
limit of 3.68 kW [85].
38
b) In Australia, the maximum generation capacity for a single-phase DG connected to a single-
phase network is 5 kW, with an export limit of 5 kW [90], except for South Australia where
a single-phase DG can have a maximum capacity of 10 kW, provided that it has an export
limit of 5 kW [106].
c) In the UK, the maximum capacity of a single-phase DG in a single-phase network is 3.68
kW. The capacity can be increased to 17 kW provided that the DG has an export limit of
3.68 kW [86], [87].
d) The maximum generation capacity for a single-phase DG connected to a single-phase
network in Turkey is 5 kW [107].
e) In Namibia, the net metering rules (NMR) require consumers to size their DG systems at
capacities up to their service connections with the DNO but the DG size must not exceed
500 kW [39]. A consumer’s service connection with the DNO is generally defined in terms
of service circuit-breaker (CB) rating or notified maximum demand (NMD) [39]. This means
that a consumer with a single-phase service connection of 60A (equivalent to 13.8 kW) is
eligible for a DG with a maximum generation capacity ≤ 13.8 kW.
Unlike other countries, Namibia has no export limit nor strict DG capacity limit. Countries
often impose export and capacity limits to prevent phase unbalance, violation of voltage
constraints, and costly upstream reinforcements, which manifest when DG penetration
levels exceed the maximum allowable penetration level – HC [88], [99].
f) A study focused on assessing the impacts of PV DG using a real distribution network
located in Virginia, USA, modeled with actual historical time-series load profile and
generation profiles estimated using historical irradiance data discovered that 25% and 55%
of consumer maximum load is the optimum penetration level for a residential and
commercial consumer, respectively [93].
g) Another study discovered that a typical urban European LV network’s aggregate DG
capacity of 70% of installed transformer capacity is generally endorsed [18]. The aggregate
capacity of 75% with respect to the transformer capacity is deemed as an acceptable DG
39
penetration level in South Africa [85]. Additionally, 15% of an MV feeder maximum load is
considered a safe penetration level for DG in South Africa and the USA, [18], [85], [91].
2.5 DG Compensation Mechanisms and Incentive Policies
Consumer-aligned renewable energy policies furnished with incentivized compensation
mechanisms such as feed-in tariffs (FiT), net metering, and net billing have played a crucial
role in the proliferation of consumer-side DG around the globe [14], [108]–[110]. These
compensation mechanisms are described as follows:
a) Feed-in tariffs (FiT) scheme
FiT is a compensation mechanism where a prosumer sells all energy generated by its
DG to the utility at a pre-defined sell rate, however, the prosumer is charged for the
energy consumed from the grid at the electricity retail rate [108]. In some parts of the
world, specifically Italy and Germany, FiT has been criticized for increasing electricity
rates, which imposed high financial burdens on electricity consumers [111]. As a result,
countries – mostly in Europe and Asia – that initially adopted FiT schemes are shifting
away, and transitioning towards net metering schemes [111], [112].
b) Net metering (NM) schemes
Net metering (NM) schemes allow prosumers to reduce their electricity bills by offsetting
a portion of their consumption with excess generations [108], [113], [114]. NM schemes
entail both net energy metering (NEM) and net billing. The term ‘net metering’ has been
used to describe both NEM and net billing, but these two compensation mechanisms
are not the same and involve different billing arrangements [112].
NEM is commonly referred to as net metering. This mechanism compensates
prosumers for excess energy supplied to the grid at the full retail rate, ratio of 1:1 [115].
Compensations are provided as energy credits in kWh. These credits are directly
applied to offset the kWh energy consumption from the grid in the current billing cycle
(e.g. one month) or can accumulate to offset the energy consumed in future billing
40
cycles. NEM attracts grid exports rather than self-consumption [10], [111], [112], [116],
[117].
Net billing is slightly different from NEM. This mechanism compensates prosumers for
excess energy supplied to the grid at an “avoided cost”, which is typically lower than the
retail tariff. Compensations are provided as a monetary bill credit in the local currency.
The energy consumed from the grid is charged at the full retail rate. At the end of the
billing cycle, the bill credit is applied directly to offset the consumption cost in the current
billing cycle or can accumulate to offset future billing cycles. As opposed to NEM, net
billing attracts self-consumption rather than grid exports [10], [14], [82], [111], [112],
[116], [117].
Net metering policies (both NEM and net billing) have been widely used to incentivize
adoptions of consumer-side DG around the world [8]. However, these policies have
been criticized for the threat they pose to traditional utility business models, especially
impacts on utility revenues, which may lead to under-recovery of fixed costs [82], [108],
[118]. To make up for these cost under-recoveries, DNOs often resort to increasing
electricity retail rates [14], [111]. This does not solve the problem but rather make net
metering policies more attractive and spur further deployment of consumer-side DG,
which eventually exacerbate the problem faced by DNOs [14], [82]. Net metering
policies also cause cross-subsidization of prosumers by non-prosumers [82], [108].
Since bill-saving benefits provided by net metering schemes depend on the structure or
design of retail tariffs. Some researchers believe that a promising solution to address
the problem faced by DNOs is to change the electricity rate design by certainly
reallocating a portion of the fixed cost recovery from the volumetric kWh charge to a
fixed charge or introduce a back-up fee or net metering charge [14], [82], [108]. Also,
the increasing consumer-side DG may lead to a shift in the timing period of peak
electricity prices; this ultimately reduces the bill-savings provided by net metering
schemes, and discourage further adoption of consumer-side DG [14].
41
2.6 Energy Storage Technologies and Applications
The deployment of renewable energy DG increases the demand for energy storage [119]. The
rising demand for energy storage is attributed to the need to smoothen the generation
fluctuations which characterize these intermittent DG e.g. wind and solar [120], [121], as well
as the need to balance the mismatch between variable electricity generation and demand by
storing surplus energy for later use [120], [121]. Energy storage does not only increase the
penetration of DG but also enhances the reliability and resilience of distribution networks by
offering applications such as peak shaving and load shifting [120]–[122]. The deployment for
energy storage is prompted by the following factors [123]:
a) Demand to improve efficiency;
b) Increasing demand to use variable renewable energy resources;
c) Rising self-consumption and self-production of energy;
d) Increasing energy access through off-grid electrification using solar PV technologies;
e) The growing emphasis on electricity grid stability, reliability, and resilience;
f) Increasing end-use sector electrification (e.g. electrification of the transport sector).
2.6.1 Types of Energy Storage Technologies
Energy storage encompasses a wide range of technologies. These technologies are classified
into five different categories based on the type of energy conversion principle employed for the
storage. These categories are mechanical, electrical, thermal, chemical, and electrochemical
energy storage technology.
2.6.1.1 Mechanical Energy Storage Technologies
These are energy storage technologies that use mechanical mechanisms and principles to
store and release energy. Mechanical energy storage encompasses flywheel, pump hydro
storage, and compressed air energy storage.
a) Flywheel energy storage
A flywheel is a mechanical device that stores energy as rotational energy using a rotating
mass. The mass is kept rotating at a constant speed. This energy is later released by
slowing down the flywheel’s rotor [123], [124].
42
b) Pumped hydro storage
This form of storage utilizes elevation changes to store electricity for later use. Electricity is
used to pump water from a lower reservoir to a reservoir at a higher elevation [123]. When
energy is needed, the water flows down and rotates turbines that generate electricity [124].
c) Compressed air energy storage (CAES)
Compressed air energy storage uses electricity to compress air and store it in storage tanks
[123]. When the energy is required the compressed air, which is a mixture of natural gas,
is released and burnt in a gas turbine to generate electricity [123], [124].
2.6.1.2 Electrical Energy Storage Technologies
These are energy storage technologies that employ electrical mechanisms and principles to
store and release energy. Electrical energy storage technologies include super-capacitors and
superconductors:
a) Supercapacitors
Super Capacitors store energy in electrostatic fields between two conductive plates
called electrodes which are separated by a small distance [43], [121], [123], [124].
b) Superconducting magnetic energy storage
This technology stores energy in a magnetic field using a superconducting coil [123],
[124].
2.6.1.3 Thermal Energy Storage Technologies
These are energy storage technologies that employ thermal mechanisms and principles to
store and release energy. Energy storage is achieved by heating or cooling a substance, e.g.
water or rocks [124]. Examples of thermal storage include ice storage and underground thermal
energy storage [123], [124].
2.6.1.4 Chemical Energy Storage Technologies
These are energy storage technologies that employ chemical mechanisms and principles to
store and release energy. According to Gardiner [124], an example of chemical energy storage
43
is the electrolysis of water into oxygen and hydrogen. The created hydrogen is stored in
pressurized containers until the energy is needed. When it is needed, the energy can be
converted back to electricity using a fuel cell.
2.6.1.5 Electrochemical Energy Storage Technologies
These are energy storage technologies that employ electrochemical mechanisms and
principles to store and release energy. Electrochemical energy storage mostly refers to
batteries, which use chemical reactions with two or more electrochemical cells to enable the
flow of electrons [3]. Some commercially available type of batteries includes the lead-acid
battery, advanced lead-acid battery, nickel-cadmium battery, sodium sulphur battery, lithium
iron phosphate, sodium nickel chloride, lithium-ion battery, etc. [120]–[122]. Lithium-ion and
lead-acid batteries are the prominent commercially available types of batteries.
2.6.2 Applications of Energy Storage Systems
Energy storage has several applications, some of which are discussed in detail below.
a) Integration of renewable energy:
Energy storage can be used to maximize the uptake of intermittent renewable energy
DG by storing energy during high generation and low demand periods and dispatching
it to the grid during high demand periods. It can also be used for smoothing renewable
generation by mitigating rapid and seasonal output changes [121]–[125].
b) Transmission and distribution congestion relief and infrastructure upgrade deferral:
Energy storage can be used to reduce energy demand during peak periods through
peak shaving. This is achieved through the concept of load shifting; by charging storage
systems from the grid during off-peak periods and discharging stored energy to the grid
during peak periods. Peak shaving reduces congestion in transmission and distribution
(T&D) networks, thus enhancing the utilization of T&D assets, and deferring
investments to upgrade or replace T&D infrastructures [121]–[125].
c) Voltage support
44
Energy storage can also be used for voltage support. This is achieved through the
injection or absorption of reactive power to maintain voltage levels in the transmission
and distribution system under normal conditions [121]–[125].
d) Frequency regulation
Continuous balancing of fluctuating power supply and demand in a power system under
normal conditions is referred to as frequency regulation [123]. Synthetic inertia is
required in order to perform frequency regulation. Synthetic inertia is the change in
power output proportional to the change in power system frequency [124]. Energy
storage can be used to add synthetic inertia to the system and subsequently perform
frequency regulation when required. The fast response of energy storage technologies
(e.g. battery and flywheels) make them favourable for frequency regulation [121]–[125].
2.7 Role of Inverter Export Limit Function to Meet Grid Code Requirements
Increasing penetration of DG in distribution networks has prompted DNOs to find mechanisms
to safeguard their networks against technical constraints associated with grid exports. In doing
so, they have tightened their grid code requirements through grid export limits. Acknowledging
the dilemma faced by DNOs, inverter manufacturers have discovered an intelligent control
mechanism to reduce the production output of an inverter without switching it off completely.
This control mechanism is referred to as an “Export Limit Function” [126], [127].
The export limit function can control the amount of power from a PV system and the power
exported to the electricity grid [126], [127]. The function requires an energy meter coupled to
the generation facility to measure the export into the grid and communicate to the inverter
through the RS485 terminal block or ethernet port on the inverter. The export limit is user-
defined and can be changed by an operator anytime. If the power flow into the grid is greater
than the predefined export limit, then the inverter will reduce its generation output dynamically
to satisfy the export limit requirements [126], [127]. To date there is no manufacturer that has
developed an inverter with the date-based multiple export limit functionality, and this makes it
very difficult to impose and define monthly grid exports limits.
DNOs in other parts of the world, through their grid code requirements, have started imposing
limits on the amount of power exported into the grid by DG. This is increasingly becoming
45
important to reduce the risk of non-compliant voltages and overload conditions on electrical
infrastructures and equipment [128]. The export limit function on inverters have made it easier
to satisfy grid code requirements by consumers because they can now change the setting of
the export limit as and when required or imposed by the DNOs [126], [127]. DNOs in Australia
and the UK have standardized export limits, especially for DG connected to single-phase
networks [88], [106], [128].
2.8 DigSilent Programming Language (DPL)
It is often cumbersome and time-consuming to investigate a large number of simulation
scenarios, especially when having parameters of data models that need to be changed for a
certain range, or results that need to be post-processed. DigSilent PowerFactory has since
introduced the concept of simulation automation using a scripting language called DigSilent
Programming Language (DPL) [129]. DPL offers a C-style scripting language with syntax quite
similar to the C++ programming language. It uses simple syntax with a versatile usage of
commands and functions, which allows it to control the PowerFactory program [129]. The
language can be used to assign external profiles or characteristics, alter model parameters,
and control simulations through scripting. Other specific applications of DPL include [129]:
a) Automation of simulation
This entails starting, repeating, and stopping a simulation. An example of this
application includes repeated execution of steady-state power flow analysis with varying
dynamic profiles, automated assignment of load characteristics, and execution of
simulation under different scenarios, which may include modelling of a different
generation and load conditions in the network [129].
b) Implementation of controls: This entails changing the system’s behaviour [129].
c) Analysis of results: This entails the evaluation and visualization of results [129].
46
3. POST-IMPLEMENTATION REVIEW OF THE NAMIBIAN
NET METERING RULES (NMR)
3.1 Provisions of the Net Metering Rules (NMR)
In 2016, the government through the Ministry of Mines and Energy (MME) and the Electricity
Control Board (ECB) developed a net metering policy, also known as the Net Metering Rules
(NMR). The NMR was developed to establish a uniform and consistent regulatory platform for
consumer-side DG, promote sustainable renewable energy sources and encourage the
adoption of consumer-side DG in the effort to diversify and increase the shares of renewable
energy in the Namibian energy mix. Despite significant contributions from other factors, the
proliferation of consumer-side DG in Namibia is mainly attributed to the NMR.
The NMR defines the minimum requirements for interconnecting consumer-side DG to the grid,
which include, inter alia, eligible generation technologies, generation capacity limits, metering
requirements, and billing and compensation arrangements [39]. The NMR requires consumers
to size DG systems at capacities up to their service connections with DNOs as long as the DG
system does not exceed 500 kW [39], as expressed in (12). A consumer’s service connection
with a DNO is generally defined in terms of service circuit-breaker (CB) rating or notified
maximum demand (NMD) [39]. The NMR also requires DNOs to process consumer-side DG
connection requests on a first-come-first-serve basis until limits imposed by stability
requirements, as determined by practical experience and technical studies are reached [39].
This means that a consumer with a single-phase service connection of 60A (equivalent to 13.8
kW) is eligible for a DG with a maximum generation capacity ≤ 13.8 kW.
PDG
≤ {P
CB/NMD, PCB/NMD < 500 kW
500 kW, PCB/NMD ≥ 500 kW (12)
Where:
- PDG
: DG system size
- PCB/NMD
: A consumer’s service connection with a DNO, which is defined in terms of
a service CB rating or NMD.
47
The NMR further mandates DNOs to compensate prosumers for excess energy exported to
the grid using net metering credits, a non-physical monetary compensation mechanism. Credits
are calculated from energy exports at a predetermined sell rate (avoided cost of energy). These
credits can be applied directly to offset the cost of the prosumer’s energy imports. However,
only export energy up to the prosumer’s import energy in the same billing period can be
converted into credits. If export energy is more than import energy, the net export energy is
carried over in cyclical to offset the cost of import energy in the next billing period subject to a
utility’s financial calendar. Credits issued and left unutilized at the end of a financial calendar
expires and are written off as free credits to the DNOs [39]. These descriptions indicate that
Namibian prosumers are compensated via ‘net billing’ despite the use of ‘net metering’ in the
NMR.
The generation capacity limit based on a prosumer’s service circuit breaker rating and the first-
come-first-serve approach to DG connection was adopted from the Interstate Renewable
Energy Council (IREC) model NMR which was developed in 2003; almost 2 decades ago [115],
[130]. Literature has indicated that many states that initially adopted practices of IREC’s model
NMR have revisited and updated their policies to incorporate lessons learned, and best
practices to help DNOs facilitate the growing DG market effectively [115].
3.2 The motivation of the NMR Review
Since the implementation of the NMR, there has been no rigorous post-implementation review
to assess its effectiveness, despite rising concerns from DNOs in Namibia and similarly DNOs
in several countries about whether NM policies are suited for long-term application in the fast-
growing market of prosumers [15]. This review conducts a broad appraisal of consumer-side
DG integration in the networks of Namibian DNOs, and an in-depth assessment of the technical
and financial impacts of the NMR using the Erongo Regional Electricity Distributor (Erongo
RED) as a case study. Based on the qualitative and quantitative findings from the study,
adaptations to current utility and regulatory policies on DG integration in Namibia are
recommended.
3.3 Choice of Case Study for the NMR Review
Distribution networks in Namibia, regardless of the operating utility, are designed to operate
within similar technical requirements as imposed by adopted standards, in this case, the NRS
48
034 [44], NRS 048 [48], and other relevant standards from SANS, IEC, and IEEE. In the same
vein, all DG behind-the-meter in Namibia are governed by the NMR. Given these common
technical aspects, the broad impacts of DG systems on distribution networks, regardless of the
operating DNO, may be comparable. Thus, the effectiveness of the NMR and its impacts on
the DNOs, consumers, and prosumers are reviewed in the context of Erongo RED.
Erongo RED pioneered DG integration in Namibia by being the first DNO to connect a DG to
its distribution network. This DG was a 220 kW wind generator, a licensed DG, connected to
Erongo RED’s distribution network in Walvis Bay in 2005 [131]. Erongo RED also has a 3.8
MW Solar photovoltaic (PV) Plant, a licensed DG, developed by an IPP through a PPA, and
connected to Erongo RED’s distribution network in Arandis. This plant was commissioned in
2016. Between 2005 and 2019, about 213 DG with a combined capacity of 10.67 MW had been
connected to Erongo RED’s grid. Two of the 213 DG with a combined capacity of 4.02 MW are
licensed DG. The remaining 211 DG, with a combined capacity of 6.65 MW, were consumer-
side DG (prosumer DG) installed for net metering (NM). Prosumer DG consists of 99 three-
phase systems on business premises with a combined capacity of 6.06 MW, and 112 single-
phase systems on residential premises with a combined capacity of 0.59 MW.
The installation statistics of prosumer DG, in terms of number and capacities installed per year
from 2011 to 2019 are shown in Fig. 12 and Fig. 13, respectively. The first prosumer DG on
Erongo RED’s grid, a PV system, was connected in 2011 via the cash-compensating FiT
program that was introduced in 2010 and abolished in 2016 after the promulgation of the NMR.
Currently, PV is the prevailing technology for prosumer DG on Erongo RED’s distribution
network. Fig. 14 displays typical import and export profiles for a prosumer with a PV system.
49
Fig. 12: Number of prosumer DG connected to Erongo RED’s grid per year
Fig. 13: Capacity of prosumer DG connected to Erongo RED’s grid per year
1
12
38 38
34
27
38
8
15
0
5
10
15
20
25
30
35
40
2011 2012 2013 2014 2015 2016 2017 2018 2019
Nu
mb
er
of p
rosu
me
r D
G
0.050.14
1.05
1.41
0.66
0.51
1.70
0.25
0.88
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2011 2012 2013 2014 2015 2016 2017 2018 2019
Ca
pa
city o
f p
rosu
me
r(M
W)
50
Fig. 14: A prosumer’s daily import/export profile
3.4 Focus and Considerations of the NMR Review
The post-implementation review of the NMR was focused on a population of 211 prosumers
registered on Erongo RED’s DG database up to the 2018/2019 financial year. This included
112 and 99 residential and business (commercial and industrial) prosumers, respectively. A
residential prosumer denotes a single-phase consumer with a DG system, while a business
prosumer denotes a three-phase consumer with a DG system. Twelve-month (01 July 2018 to
30 June 2019) energy import and export datasets for the selected prosumers were downloaded
from Erongo RED’s automatic meter reading (AMR) system and analyzed. The analysis
included:
a) Determining the magnitude, date (month and day of the week), and time of the
maximum instantaneous power (kW) exported to the grid by each prosumer’s DG
system during the review period. The maximum instantaneous power (kW) exported to
the grid is subsequently denoted as ‘peak export power’.
b) Classification and grouping of prosumers according to the date and time at which their
peak export powers were recorded to examine the criticality of energy export to the grid.
c) Determining the service circuit-breaker ratings of residential prosumers and NMD of
business prosumers and the approved sizes (capacities) of DG systems for individual
prosumers.
d) Review of the distribution network planning and design standards to determine the
maximum ADMDs available for residential distribution network designs.
0
2
4
6
8
10
Imp
ort
and
exp
ort
pro
file
(kW
)
Grid Energy Export Grid Energy Import
51
e) Comparison of prosumers’ NMD or service circuit-breaker rating with the approved DG
system capacity to determine compliance with the NMR.
f) Comparison of prosumers’ approved DG system capacity to prosumers’ peak export
power to determine the generation capacity factor at which their DG systems exported
maximum instantaneous power to the grid.
g) Comparison of residential prosumers’ approved DG system capacity to the maximum
recommended ADMD for residential distribution network designs.
h) Determining prosumers’ total energy imports and exports.
i) Calculation of the monetary value of energy imports (import cost) and energy exports
(net metering credits) for each prosumer. Erongo RED’s retail and sell rates (avoided
cost of energy) for the 2018/2019 financial year were used for these calculations. The
electricity rates were converted to US dollars (USD) using the average exchange rate
of 14.45 NAD/USD for 2019 [132]. For residential prosumers, a retail tariff of 0.143 USD
/ kWh and a sell rate of 0.084 USD / kWh were used to calculate the monetary value of
energy imports and exports, respectively. For business prosumers, an average retail
tariff (0.139 USD/kWh) and sell rate (0.081 USD/kWh) were determined from the time
of use (ToU) import and export tariffs, respectively (Table 8), and used to calculate the
monetary value of energy imports and exports.
j) Determining the proportion of prosumers’ energy import costs offset by their net
metering credits and Erongo RED’s revenue losses due to net metering compensations.
Table 8: Business (Three-Phase) tariffs [133]
ToU tariff category (Low
season)
Peak tariff
(USD/kWh)
Standard tariff (USD/kWh)
Off-Peak tariff (USD/kWh)
Average tariff
(USD/kWh)
Import 0.163 0.143 0.110 0.139
Export (Small
Renewable) 0.105 0.084 0.053 0.081
52
3.5 Results and discussions
The total installed DG capacities of the residential and business prosumers reviewed is 0.59
MW and 6.06 MW, respectively. The results of the technical and financial analyses of both
prosumer groups are discussed subsequently.
3.5.1 Technical analysis results
The results of the technical analysis are summarized in Fig. 15, Fig. 16, and Fig. 17.
Participation in energy exports to the grid is higher among residential prosumers (97 %
participation) than business prosumers (91.92 % participation). Peak power exports for over 50
% residential prosumers and 33 % business prosumers occurred between 12:30 - 14:00, at
generation scaling factors of 0.7 to 1, i.e. over 70 % of the energy generated by these
prosumers at the time when their instantaneous peak export powers were recorded was
exported to Erongo RED’s grid (Fig. 16). This implies that residential vis-a-vis business
prosumers are characterized by low demand during the peak generation hours of their DG. In
a typical week, residential prosumers are more likely to export energy to the grid on
Wednesdays and Sundays (Fig. 17). For business prosumers, the likelihood of energy exports
is highest on Sundays, but relatively low on weekdays (Fig. 17). This is quite expected as
business activity levels (hence, energy consumption) is generally higher during weekdays than
weekends. Overall, summer/spring months (January-March and September-December) are
characterized by high participation in energy exports by both prosumer groups due to the higher
PV output from high solar irradiation levels (Fig. 15). This time-based characterization of the
participation in energy exports by prosumers is very useful for identifying when prompt
intervention measures are required to maintain grid stability.
Residential prosumers were fully compliant with the NMR regarding generation capacity limits.
However, over 56% of residential DG systems were sized beyond the maximum recommended
ADMD for residential distribution network designs of 4.5 kVA. This does not violate any code
but owing to low demands associated with residential distribution networks in Namibia, it could
increase the risk of multiple prosumers simultaneously generating and exporting excess power
to the grid, with peaks that are greater than a distribution network’s design capacity risking
network voltage rise and unbalance.
53
Non-compliance with the NMR was observed among some business prosumers; about 24.24
% of business prosumers violated the rules on generation capacity limits. About 15.15 % of the
business prosumers had NMDs that were lower than the capacities of their DG systems. In
Erongo RED’s context, business consumers are required to declare NMDs annually for the
calculation of network charges. It is plausible that some prosumers exploited a loophole in
annual NMD declarations that allowed them to avoid high network charges by declaring lower
NMDs. Thus, these prosumers pay less for network charges, shifting the cost to other
consumers. Further, 9.09 % of business prosumers recorded peak export powers beyond their
respective approved DG system capacities. This indicates that these prosumers expanded their
generation capacities without consent from Erongo RED and without simultaneously expanding
their NMDs. They did so to avoid declaring a high NMD, which is required to qualify for bigger
DG system capacities, hence avoiding high network charges.
Although it is broadly perceived that having multiple DG-behind-the-meter improves reliability
indices – system average interruption duration index (SAIDI), system average interruption
frequency index (SAIFI), customer average interruption duration index (CAIDI), and customer,
average interruption duration index (CAIFI) – through the reduction of sustained consumer
interruptions from generation inadequacy or network faults, excess power exports could cause
other undesirable technical constraints including overvoltage and equipment overload. Where
these constraints lead to equipment or line outages, consumers without DG may be the most
susceptible to interruption of supply and incursion of interruption costs. These impacts may be
negligible when the concentration of DG on the same distribution network is low. However, as
the DG concept in Namibia advances and the number of grid-integrated DG increases, these
impacts may be immense, especially if violations of the NMR are unchecked.
54
Fig. 15: Monthly distribution of the occurrence of prosumers’ annual peak export power
Fig. 16: Hourly distribution of the occurrence of prosumers’ annual peak export power
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
14.00%
16.00%
18.00%
Pe
rce
nta
ge
of p
rosu
me
rs (
%)
Residential Prosumers Business Prosumers
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
07:3
0a
m
08:0
0a
m
08:3
0a
m
09:0
0a
m
09:3
0am
10:0
0a
m
10:3
0a
m
11:0
0a
m
11:3
0a
m
12:0
0p
m
12:3
0p
m
13:0
0p
m
13:3
0p
m
14:0
0p
m
14:3
0p
m
15:0
0p
m
15:3
0p
m
16:0
0pm
16:3
0p
m
No
Exp
ort
Pe
rce
nta
ge
of p
rosu
me
rs (
%)
Residential Prosumers Business Prosumers
55
Fig. 17: Daily distribution of the occurrence of prosumers’ annual peak export power
3.5.2 Financial analysis results
The results of the financial analysis are summarized in Fig. 18 and Fig. 19. The results show
that NM compensations subject Erongo RED to energy revenue losses of approximately USD
26, 759.94, and USD 82, 883.16, which translates to 25.01 % and 2.91 % of the total monetary
value of energy imports from the grid by residential and business prosumers, respectively. This
loss is the difference between utilized net metering credits, which benefits the prosumer, and
unutilized net metering credits, which benefits the DNO. Energy revenues from residential
prosumers are impacted by NM compensations far more than revenues from business
prosumers. Revenue losses due to NM compensations add to the unquantified revenue loss
emanating from the reduced on-site energy requirement from the grid by the prosumers.
Even though NM compensation is perceived as a revenue loss to the DNO in the context of
reduced income, seeing that grid exports in Namibia are credited at the ‘avoided cost’ of the
DNO, the income lost due to NM compensations is balanced out by the reduction in the DNO’s
energy cost from the national utility (NamPower). The avoided cost at which grid exports are
credited excludes other energy service charges i.e. reliability charge, transmission losses
charge etc., which are generally charged on top of the ‘avoided cost’ if the same energy was
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
Pe
rce
nta
ge
of p
rosu
me
rs (
%)
Residential Prosumers Business Prosumers
56
procured from the national utility. The exclusion of other energy service charges on the credits
makes grid exports from prosumers a cheaper source of energy to the DNO as compared to
DNO getting the same energy from the national utility. This then nullifies NM compensations
as total loss to the DNOs, and hence leaving revenue lost due to prosumers’ reduced on-site
energy requirement from the grid by offsetting demand with generation in real-time as the only
and inevitable loss incurred by DNOs from self-generation practices.
Fig. 18: Residential prosumers (energy import value vs energy export value)
Fig. 19: Business Prosumers (energy import value vs energy export value)
73,022
33,954
106,976
7,194
0 30,000 60,000 90,000 120,000
Reduced/Net Energy Sales Revenue(Imports)
Net Metering (NM) Credits(Exports)
Gross Energy Sales Revenue(Imports)
Monetary Value(USD)
Utilized NM credits
Unutilized NM credits
2,743,433
105,003
2,848,436
22,120
0 1,000,000 2,000,000 3,000,000
Reduced/Net Energy Sales Revenue(Imports)
Net Metering (NM) Credits(Exports)
Gross Energy Sales Revenue(Imports)
Monetary Value(USD)
Unutilized NM credits
Utilized NM credits
57
3.6 Highlights of the review
Consumer-aligned renewable energy incentives have played a crucial role in the deployment
of prosumer DG in Namibia. The status quo shows that prosumer DG are expected to export
large volumes of energy to distribution networks between 12:00 - 14:30 during non-winter
months, with Wednesdays and Sundays identified as the critical days of the week. Based on
the current NMR, these exports bring financial benefits to both prosumers, and DNOs when
compensated at the avoided cost as prescribed by the NMR. Unutilized net metering credits
(energy obtained at zero cost) also add to the benefits of DNOs from grid exports.
Unlike in countries where grid exports are credited at a ratio of 1:1 leading to DNOs completely
losing revenues as a result of NM compensations, in Namibia grid exports are credited at the
avoided cost, which is generally lower than the DNOs retail tariff and equal the cost of energy
from the national utility exclusive of other energy service charges. This makes grid exports a
cheaper alternative source of energy to the DNO as compared to getting the same energy from
the national utility (NamPower), which generally adds on top of the avoided cost other energy
service charges i.e. reliability charge and transmission losses etc. Despite the benefit to DNOs,
large volumes of grid exports may however pose negative technical constraints on the grid,
which may be costly to mitigate.
DNOs are however expected to lose considerable revenues not because of NM
compensations, but because of reduced volumes of energy sales caused by the reduction in
energy consumption from the grid by both residential and business prosumers when offsetting
demand with generation in real-time. Additional revenue loss is expected as a result of peak
shaving among business prosumers, whose fixed charges are linked to demand peaks.
It is worth restating that several countries that initially adopted IREC’s model NMR have
revisited and updated their policies to incorporate the lessons learned and best practices to
help their DNOs facilitate the growing DG market effectively. The results of the analysis in this
study and careful review of best practices on the grid-integration of DG systems around the
world highlighted areas for revising the policies on the grid-integration of DG in Namibia and
similar countries. Policymakers may adopt the suggested revisions in this study to improve
current practices on the grid-integration of DG systems and sustainably drive their deployment.
58
3.7 Policy imperatives
The review of Namibia’s NMR led to an identification of important areas for policy
interventions.
a. Using a consumer’s circuit breaker rating to determine the DG capacity limit, without an
effective technical control mechanism for grid exports, may compromise the reliability
and power quality of distribution networks due to voltage magnitude rise and voltage
unbalance in areas supplied using single-phase installations.
b. The capacity dissemination methodology based on the “first-come-first-serve” basis
does not acknowledge the increasing demand for grid access to interconnect DGs,
especially in business areas and middle/high income residential areas where self-
generation is becoming lucrative as a result of energy cost savings and peak shaving.
It also does not acknowledge the long-term existence of DG, even at the time when DG
technologies (including storage) may become integral components of future power
systems. Further, this methodology favors a few individuals who can afford DG in the
present time but disregard consumers that may not afford it in the present time, though
looking forward to installing it in the future. This makes the methodology unfair and
discriminatory, thus raising inclusivity concerns in the long run.
c. The current generation capacity limit and capacity dissemination methodology may
collectively cause a fast depletion of the DG hosting capacity of distribution networks
and limit the number of DG market participants without consideration for future
participants. This makes the NMR unsustainable looking at the horizon and evolving
renewable energy markets.
To facilitate the increasing demand for access to distribution networks for DG deployment
effectively, the NMR needs to be revised from a technical and regulatory perspective to expand
the DG market platform to accommodate many participants. Accordingly, the following
adaptations are suggested:
a. Current and prospective prosumers should be accommodated in a fair, inclusive, and
sustainable manner, especially in areas where high uptake of DGs is expected such as
business areas and middle/high income residential areas. This can be achieved either
59
by setting individual generation capacity limits within the design capacities of distribution
networks, as in the case of South Africa or by imposing grid export restrictions, as in
the case of Australia and the UK. The latter is more favorable because it gives
prosumers the flexibility to optimize their electricity use by matching their generation to
their demands and manage their excess generations through either energy storage or
power curtailment. Violations of export restrictions may be penalized with a graduated
charge scheme for power exports exceeding a set threshold.
b. The expanded DG market will prepare the DNOs for interactive and competitive
markets, which offer efficient wholesale market operations, enhanced reliability, and
minimum infrastructure congestion and constraints. Competitive markets will motivate
DNOs to transition away from traditional business models or revise them to incorporate
consumer and market-driven transactive energy models. This includes grid
modernization to enable them to sell services beyond energy, with emphasis on
ancillary services and facilitation of peer-to-peer energy transactions. Competitive
markets will also necessitate a redesign of traditional tariff structures to incorporate the
presence of DG and eliminate cross-subsidization of prosumers by normal consumers.
c. There is a great need for DNOs and regulators in Sub-Saharan Africa to conduct
technical studies to determine the maximum amounts of DG that could be integrated
into their distribution networks without causing adverse technical impacts. With a
growing demand for DG deployment, these technical studies should assume the
likelihood of all consumers becoming prosumers to emulate the future electricity market
in which DG technologies will become integral components of future power systems.
d. DNOs must change their electricity rate design by certainly reallocating a portion of the
fixed cost recovery from the volumetric kWh charge to a fixed charge.
e. The Electrify Control Board (ECB), the electricity regulator in Namibia and the custodian
of the net metering policy, needs to acknowledge the risk of funds inadequacy faced by
DNOs to maintain electrical infrastructures and mitigate technical constraints caused by
consumer-side DG. This must be reflected in the policy by defining generation capacity
limits and export limits to ensure that consumer-side DG is within network hosting
capacities.
60
4. DEVELOPMENT OF METHODOLOGY FOR ASSESSING
HOSTING CAPACITIES (HC) IN EXISTING LV
DISTRIBUTION NETWORKS
This study proposes and develops a methodology for assessing hosting capacities of
consumer-side PV DG in existing LV distribution networks. The methodology adopts
techniques from hosting capacity methodologies discussed in [20], [93], [94], [97], [100], [102],
[104], [105]
4.1 The motivation of HC Methodology
This methodology is proposed as an adaption to the current consumer-side DG regulatory
policy in Namibia, known as the net metering rules (NMR). Motivations against the NMR
include:
a) From a technical view, the DG sizing methodology based on consumer’s service
connection as recommended by the NMR may limit DG participation to a few consumers
due to network capacity bottlenecks caused by the design philosophy of distribution
networks, which is centered on diversity.
b) The NMR poses the risk of causing technical constraints in distribution networks
because hosting capacities were not known before policy development.
c) The ‘first-come-first-serve’ approach recommended by the NMR lacks inclusivity,
especially in areas experiencing high and uniform uptake of DG, this includes business
areas and middle-high income residential areas. DNOs therefore need an approach that
has the potential to consider all consumers as prospective prosumers and reserve HC
capacities accordingly.
d) With increasing consumer preference for self-generation and growing demand for
access to distribution networks for DG deployment, DNOs need cost-effective
mechanisms to distribute and allocate hosting capacities to prosumers in a fair,
61
inclusive, and sustainable manner to ensure safe and reliable operation of distribution
networks upon DG integration, without requiring costly network reinforcements.
e) A long-term solution to help DNOs manage DG integration is required. This is because
advances toward smart-grids consider DG systems as integral components of future
power systems.
A new methodology is therefore proposed and developed to help DNOs, facing high uptake of
DGs in distribution networks in Namibia and other countries in Africa, evaluate the HC of their
distribution networks. This methodology is however limited to existing distribution networks in
areas where a high and uniform uptake of DGs is expected. These areas include business
areas and middle/high income residential areas, where most DG integration requests are
received. High demand for DG in these areas signals the likelihood of every consumer
becoming a prosumer.
This algorithm can be considered as an improved version of traditional deterministic methods
because the time-series load and generation data have eliminated the worst case, single
scenario modelling. It can also be considered as a probabilistic method because time-series
load data obtained from smart meters and time-series generation profiles predicted using local
solar irradiance data are stochastic and variable. DG location uncertainty is addressed because
DGs are located at consumer’s premises. The issue of generation capacity uncertainty is also
addressed by allocating each consumer a DG with a maximum capacity up to the service
connection with the DNO as recommend by the NMR. The issue of load and generation
uncertainty is also addressed because the algorithm considers time-series historical load data
and generation profiles forecasted using historical solar irradiation data. This helps the
algorithm to capture time correlations between demand and generation.
4.2 Considerations of HC Methodology
As adopted from HC methodologies presented in the literatures, this methodology requires a
distribution network model with real network parameters, loads and PV systems [20], [93], [94],
[97], [104], [105]. Loads require time-series load profiles with a time resolution of 30-minutes
for a period of 1 to 12 months. A PV system requires a generation profile estimated from the
62
PV system’s nominal capacity and historical solar irradiance data of distribution network’s
locations.
The methodology introduces a “Generation Scaling Factor (GSF)” ranging from 0% - 100%,
which increments with a step-size of 1%. The GSF modifies the nominal capacities of PV
systems uniformly. The methodology further features iterative load flow simulations (a load flow
at each GSF) facilitated by an extensive search algorithm that evaluates network parameters
identified as performance indices (voltage and loading of lines and transformers) against
predefined performance limits at each iteration to find the HC of the network. The methodology
produces HC at a consumer level and network level.
Uniquely, the methodology introduces a concept of calculating HC on a monthly basis. This
concept eliminates the practice used in similar methodologies focused on calculation of HC
based on a single extreme event encountered over an entire e.g. 12 months. Instead, the
proposed concept calculates HC on a monthly basis, based on extreme events encountered
monthly. This approach acknowledges the influence of monthly weather variations on load
consumption and solar irradiation, which have a direct impact on HC. Since calculations are
carried out monthly, if a period of 12 months is considered, the methodology would produce 12
consumer HC results and 12 network HC results. The bottom line of the methodology is that it
captures uncertainties in PV generation and time dependency correlations between load and
generation profiles. This methodology also acknowledges the presence of existing DG systems
on distribution networks.
The methodology uses the modeling and simulation platform in DigSilent PowerFactory
employing the quasi-dynamic simulation tool. The search algorithm featured in the
methodology utilizes a script developed using DigSilent Programing Language (DPL).
4.2.1 Modelling Approach
Fig. 20 illustrates the general modeling approach of the methodology, which involves the
modelling of a distribution network as per as-built drawing, addition of well-defined loads and
PV systems, and automated simulations using the DPL script.
63
Fig. 20: General modeling approach
4.2.2 Distribution Network Model
Build a distribution network model in DigSilent PowerFactory to resemble the as-built drawing
structures of the distribution network under assessment. The phase configuration of the loads
should be considered (whether three-phase or single-phase). Parameters of key network
equipment such as transformers and cables should be defined accurately as per the as-built
drawing. All consumers on the network are assumed as prosumers. The network’s standard
nominal low voltage in Namibia is 400 V three-phase, or 230 V single-phase, as recommended
by design standards [44], [48]. Balanced load flow simulation should be considered for
balanced three-phase networks, while unbalanced load flow simulation should be considered
for unbalanced single-phase networks.
4.2.3 Consumer Load Model
Each consumer load is defined using a customer-specific historical time-series load data in the
intervals (time resolution) of 30 minutes and covering a period of 1 to 12 months. The load data
64
is obtained from a consumer-specific smart meter. Loads are set to operate at a power factor
of 0.99. Three-phase consumers are modelled as three-phase balanced loads, while single-
phase consumers are modelled as single-phase unbalanced loads to capture their inherent
unbalanced behaviors.
The proposed methodology acknowledges the presence of existing DG systems on distribution
networks. In the case where the network already have existing prosumers, the methodology
will evaluate the available remaining HC taking cognizant that existing DG systems have
already absorbed a portion of it. DNO’s DG databases should be used to identify existing
prosumers on networks so that they can be modelled as positive loads and negative loads to
reflect energy consumed from the gird and energy exported to the grid respectively. Smart
meters store energy imports and energy exports in different registers and this makes it easier
to differentiate them.
4.2.4 PV System Model
Similar to loads, PV systems are modelled as either three-phase or single-phase systems
depending on the corresponding consumer load type. PV systems are set to operate in the
‘solar calculation’ mode, which activates the calculation of generation profiles according to
historical solar irradiance data of the network site. This is achieved by defining the geographical
coordinates of the site in DigSilent PowerFactory. PV systems are further set to operate at a
constant power factor of one. Capacity and technology of PV panels and inverters are also
defined. In the context of this study, a 380W monocrystalline PV panel and inverters sized
according to consumer’s NMD are considered in the modelling of PV systems. The generation
profile of a PV system is governed by the PV system’s nominal capacity, efficiency of PV panel
technology, local solar irradiance data (determined by the Haurwitz model), and a generation
scaling factor (GSF), which modifies a PV system’s nominal capacity proportionally. The
Haurwitz model for global horizontal irradiance (GHI) is available in DigSilent PowerFactory
[134]–[136]. A GSF in the range of 0% to 100% (incrementing by 1%) is applied uniformly to all
PV systems in the model, and a load flow is performed at each GSF until performance metrics
(voltage limits and thermal limits) are violated. The mathematical expression shown in (13)
illustrates how the above mentioned factors govern the output of a PV system as a function of
time.
65
PVoutput
(t) = γGSF*GH(t)* PVCapacity ∗ 𝜂
(13)
Where:
- PVoutput
: Instantaneous PV system generation output in kW.
- γGSF: Generation Scaling Factor (GSF) from 0%-100%, with an incremental value of 1%.
- GH(t)=1098* [cos(∅z(t)) *e-0.057
cos(∅z(t))], GH(t) is the Haurwitz model for global horizontal
irradiance as a function of time. ∅z(t) is the zenith angle as a function of time.
- PVCapacity
: The nominal capacity of the PV system in kW.
- η: Efficiency of the PV technology employed
Existing prosumers are modelled as follow:
a) If a prosumer’s existing PV capacity is equal to the prosumer’s NMD then no new PV
system will be modelled for this prosumer. However, a positive load (denoting
consumption from the grid) and a negative load (denoting export to the grid), as
obtained from a smart meter, will be modelled as the prosumer’s loads.
b) If a prosumer’s existing PV capacity is less than the prosumer’s NMD, then a new PV
system with a capacity equals the difference between the prosumer’s NMD and the
prosumer’s existing PV capacity will be modelled for this specific prosumer, in addition
to the positive and the negative load.
4.2.5 Performance Indices
The methodology features iterative load flow simulations (a load flow at each GSF) facilitated
by an extensive search algorithm that evaluates network parameters identified as performance
indices (voltage and loading of lines & transformers) against predefined performance limits at
each iteration to find the hosting capacity (HC). The performance indices and limits, which are
presented in Table 9, were considered in the hosting capacity methodology.
66
Table 9: Performance indices and limits
Performance indices Limits
Bus Voltage 1.1 pu
Line Loading 100%
Transformer Loading 100%
4.3 Proposed Methodology
A search and simulation algorithm shown in Fig. 21 defines the pseudocodes and processes
involved in the coding and development of a script-based methodology for assessing hosting
capacity in LV distribution networks, as discussed in section 4.1 and 4.2. The methodology was
developed using DigSilent Programming Language (DPL). The script models different
generation scenarios under dynamic load and uncertain generation conditions to assess the
maximum amount of DG that can be accommodated in a network without adverse voltage and
loading impacts.
The script’s core objective is to assess the DG HC of distribution networks and subsequently
produce HC results at a consumer level as expressed in (14), and network level, as expressed
in (15). The script additionally prints out the results to the output window of DigSilent
PowerFactory and exports the plotted results, such as voltage and loading profiles as WMF
pictures to a predefined result folder in a computer’s C Drive. The script is amenable to
modifications, especially when dealing with data having a duration of less than or more than a
year.
HCConsumer(%) =
PVOptimum Size
PCB/NMD *100% = γGSF (optimum)
(14)
Where:
- HCConsumer(%): Consumer hosting capacity in percentage.
- PVOptimum Size
: Optimum PV capacity/size.
- PCB/NMD
: Consumer’s service connection with utility.
67
- γGSF (optimum) : The generation scaling factor (GSF) between 0% and 100% where
violation of performance indices first occurs.
HCNetwork(%) =
TrLoad at γGSF (optimum)
TrRating *100%
(15)
Where:
- HCNetwork(%): Network hosting capacity in percentage.
- TrLoad at γGSF (optimum)
: Load (kVA) on the Transformer at γGSF (optimum).
- TrRating
: Transformer capacity rating (kVA).
68
Fig. 21: Modeling and hosting capacity assessment algorithm.
69
5. TESTING AND RESULTS OF METHODOLOGY FOR
ASSESSING HOSTING CAPACITIES
The developed HC assessment methodology was tested on one residential and one business
LV distribution network, part of Erongo RED’s distribution network in Walvis Bay, Namibia. The
objective of the test was to quantify the effectiveness of the methodology and discuss results
obtained. Three tests were conducted in total (two on !Nara residential distribution network and
one on Spar business distribution network).
5.1 Summary of the Global Horizontal Irradiance (GHI) of Walvis Bay
A summary of the monthly global horizontal irradiance (GHI) for Walvis Bay, where both !Nara
network and Spar network, are located is presented in Fig. 22. This summary was obtained
from HelioScope, a web-based PV design software [137]. Solar irradiation data is useful in the
interpretation of the HC results.
Fig. 22: Monthly global horizontal irradiance (GHI) for Walvis Bay
0
50
100
150
200
250
Glo
bal
Ho
rizo
nta
l Irr
adia
nce
(G
HI)
kWh
/m2
70
5.2 Testing HC Methodology on !Nara Residential Distribution Network (Test 1)
The HC assessment methodology was tested on a residential distribution network called !Nara
distribution network, which supplies power to 95 residential consumers. At the time of data
collection, this network did not have any prosumer.
5.2.1 As-Built Drawing of !Nara Network
The !Nara low-voltage (LV) network is supplied from a secondary substation called !Nara
substation, which contains a 315 kVA (11/0.42 kV) transformer. The network contains ten
distribution kiosks on three LV circuits. Fig. 23 shows the cadastral layout of !Nara distribution
network. Fig. 24 shows the network configuration, summarizes the number of kiosks and
connected consumers, and provides properties of LV cables interconnecting kiosks
downstream the substation. In this test, consumers were modelled as single-phase
(unbalanced) loads to demonstrated the reality on the ground. A summary of assumed
distribution of single-phase consumers on individual phases at respective kiosks is shown in
Table 10.
Fig. 23: Cadastral layout of !Nara network
71
Fig. 24: !Nara distribution network configuration and properties
Table 10: Distribution of single-phase loads on phases
Kiosk
Phase
K
1.1
K
1.2
K
2.1
K
2.2-1
K
2.2-2
K
2.3
K
3.1
K
3.2
K
3.3
K
3.4
Total
Per
Phase
Phase 1
(Red) 4 3 5 3 3 3 3 3 3 2 32
Phase 2
(White) 3 4 4 2 4 4 3 3 3 2 32
Phase 3
(Blue) 3 3 5 2 3 3 4 3 3 2 31
Total per kiosk 10 10 14 7 10 10 10 9 9 6 95
72
5.2.2 DigSilent Model of !Nara Network in DigSilent PowerFactory 2018
!Nara network was modeled in DigSilent PowerFactory 2018 according to the as-built drawing
received from Erongo RED as shown in section 5.2.1. The model consists of an MV/LV
transformer, LV feeder cables, distribution kiosks and consumer loads. A summary of assumed
distribution of single-phase consumers on individual phases at respective kiosks is shown in
Table 10.
The standard service connection with any utility in Namibia for a residential consumer is a
single-phase circuit breaker rating of 60A [138], which is equivalent to 13.8 kVA. According to
the Namibian net metering policy, a residential consumer can put up a PV system with a
maximum capacity of 13.8 kW. Each consumer in the model was therefore allocated a PV
system with a nominal capacity of 13.8 kW. The loads and PV systems were modelled and
defined as described in Chapter 4. Historical time-series metering data (with a time resolution
of 30 minutes) for individual consumers on !Nara’s network covering a period of 12 months
(July 2018 - June 2019) were considered in the definition of the loads. The data was
downloaded from Erongo RED’s automatic meter reading (AMR) system. The geographical
coordinates of !Nara network were defined in DigSilent to provide the software access to
meteorological data of this location.
The DigSilent model of !Nara network is shown in Fig. 25. Fig. 26 shows a typical outline of a
distribution kiosk, illustrating how consumer loads and PV systems are connected in the model.
73
Fig. 25: DigSilent network model for !Nara network
74
Fig. 26: Inside configuration of a distribution kiosk model
5.2.3 Simulation Results and Discussion of the HC results of !Nara Network (Test 1)
!Nara network model was simulated using the developed DPL script to produce results such as
voltage profiles; loading profiles, and script summary reports highlighting HC results on a
monthly basis.
a) HC results for !Nara Network (Test 1)
Fig. 27 below shows a typical monthly script summary report indicating the HC results for June
2019. The script summary reports for the remaining months are provided in APPENDIX A. The
typical loading profile and voltage profile of !Nara network during an event of violated
performance metric is provided in APPENDIX C, with reference to June 2019. Voltage and
loading profiles for the rest of the months have been submitted with this report as
supplementary information.
Fig. 28 presents HC results of !Nara network at a consumer level, and network level,
respectively. Table 11 shows a summary of the types, and time of occurrence of technical
constraints encountered during the calculation of the HC results, which was derived from script
summary reports, voltage profiles, and loading profiles.
Fig. 27: Script summary report for !Nara network during June 2019 (Test 1)
75
Fig. 28: Network and Consumer level monthly HC results for !Nara network (Test 1)
Table 11: Technical constraints encountered on !Nara network (Test 1)
Month Network constraint Time
July 2018 Overvoltage on phase B 11:19
Line Overload 11:30
August 2018 Overvoltage on phase C 13:00
September 2018 Line Overload 15.54
October 2018 Overvoltage on phase A & B 12:27
Overvoltage on phase A 12:23
November 2018 Overvoltage on phase C 12:28
Overvoltage on phase A 12:56
December 2018 Overvoltage on phase C 13:24
Overvoltage on phase B 11:56
January 2019 Line Overload 09:30
Overvoltage on phase C 09:19, 09:28 & 09:30
February 2019
Overvoltage on phase A 12:24, 12:28
Overvoltage on phase A, B & C 13:25
Overvoltage on phase C 11:58, 13:21
March 2019 Overvoltage on phase C 11:59
April 2019 Overvoltage on phase A 13:46 & 14:00
0%
10%
20%
30%
40%
50%
60%
70%
80%
0%
5%
10%
15%
20%
25%
Per
cen
tage
of
tran
sfo
rmer
rat
ing
%
Per
cen
tage
of
uti
lity
serv
ice
con
nec
tio
n(C
B o
r N
MD
) %
Monthly HC at Consumer Level Monthly HC at Network Level
76
Overvoltage on phase A & B 13:36
Line Overload 13:43
May 2019 Overvoltage on phase C 12:28
June 2019 Line Overload 12:49
Overvoltage on phase C 12:38 & 13:00
b) Discussions of HC results for !Nara Network (Test 1)
Consumer HC results obtained in Test 1, as shown in Fig. 28, varied monthly from 10% to 20%.
These percentages correspond to 1.38 kW and 2.76 kW, respectively. August had the lowest
consumer HC of 10%, while April had the highest consumer HC of 20%. Results in Fig. 28
shows that HC at a network level ranged from 44% to 74% of the substation transformer rating.
Both consumer and network monthly HC results of !Nara network followed a similar pattern of
low HC in months of low solar irradiation (Winter months), and high HC in months of high solar
irradiations (summer months). The HC pattern was directly proportional to Walvis Bay’s monthly
irradiance pattern, shown in Fig. 22.
Table 11 show that technical constraints, mainly overvoltage and line overload, were
encountered between 09:00 and 15:00, the time during which high solar irradiance levels are
generally recorded in Namibia. Overvoltage was recorded frequently in the test as compared
to line overload and this demonstrate the dominance of overvoltage as a technical constraint
in residential distribution networks. Overvoltage conditions did not develop on all three phases
concurrently but rather developed on individual phases. This phenomenon is attributed to
phase imbalance, and confirms phase imbalance as a limiting factor of HC in single-phase
networks. Consumer HC results obtained in this test were lower than limits employed in HC
practices of countries reviewed in this study and HC results obtained in similar studies for
similar type of networks.
The HC results obtained in this study necessitates two innovation opportunities for inverter
manufacturers. The two innovation opportunities have the potential help most countries abolish
restrictive DG sizing measures, which limits the uptake of DGs in distribution networks, to
provide prosumers the flexibility to have optimally sized DGs to meet their energy requirements
while keeping the networks safe. These two innovation opportunities are described below:
77
- Seeing that technical constraints are encountered frequently during hours of high solar
irradiation, it would add value if inverter manufacturers can develop inverters that
automatically switches into ‘load following mode’ (produce according to on-site demand)
or automatically block exports to the grid during hours of high solar irradiation to
minimize technical constraints associated with grid exports. This will give prosumers
the flexibility to optimize and match their generation to their demands. Or,
- Seeing that HC results vary monthly, it would also add value if inverter manufacturers
can develop inverters with a date-based multiple export limit functionality. This function
will allow utilities to impose monthly export limits according to monthly HC limits
obtained from technical studies to minimize technical constraints, while giving
prosumers the flexibility to optimize and match their generation to their demands. DNOs
can communicate these limits to prospect prosumers during approval process of DGs
and impose them during commissioning of DG installations.
5.3 Testing HC Methodology on !Nara Residential Distribution Network (Test 2)
The HC methodology was retested on !Nara distribution network with loads and PV systems
modelled as three-phase balanced systems. The objective of the retest was to analyze and
understand how the results compares to HC results of an unbalanced system tested in Test 1
to draw relevant conclusions on the influence of phase unbalance on HC results and derive
appropriate recommendations that can benefit future researchers in the same area. The
complete modelling of !Nara’s network is as described in section 5.2, the only difference is that
this test considered phase balancing of the loads and PV systems by modelling and simulating
the network as three-phase balanced system instead of single-phase unbalanced system.
5.3.1 Results (Test 2)
Fig. 29 below shows a typical monthly script summary report indicating the HC results for June
2019. The script summary reports for the remaining months are provided in APPENDIX B. The
typical loading profile and voltage profile of !Nara network during an event of violated
performance metric is provided in APPENDIX D. Voltage and loading profiles for the rest of the
months have been submitted with this report as supplementary information.
78
Fig. 30 presents HC results of !Nara network at a consumer level, and network level,
respectively. Table 12 shows a summary of the type and time of occurrence of technical
constraints encountered during the calculation of the HC results, which was derived from script
summary reports, voltage profiles, and loading profiles.
Fig. 29: Script summary report for !Nara network during June 2019 (Test 2)
Fig. 30: Network and Consumer level monthly HC results for !Nara network (Test 2)
Table 12: Technical constraints encountered on !Nara network (Test 2)
Month Type of network
constraint Time of occurrence
July 2018 Overvoltage 13:00
August 2018 Overvoltage 12:30, 13:00, 13:30
September 2018 Overvoltage 12:30
October 2018 Overvoltage 13:00
0%
10%
20%
30%
40%
50%
60%
70%
80%
0%
5%
10%
15%
20%
25%
30%
35%
Per
cen
tage
of
tran
sfo
rmer
rat
ing
(%)
Per
cen
tage
of
uti
lity
serv
ice
con
nec
tio
n(C
B o
r N
MD
) %
Monthly HC at Consumer Level Monthly HC at Network Level
79
November 2018 Overvoltage 11:30, 12:00, 12:30, 13:00, 13:30
December 2018 Overvoltage 12:00, 12:30, 13:30
January 2019 Overvoltage 13:00, 13:30, 14:00
February 2019 Overvoltage 12:30, 13:00, 13:30, 14:00, 14:30,
15:00
March 2019 Overvoltage 13:00, 14:00
April 2019 Overvoltage 13:00, 13:30
May 2019 Overvoltage 13:30
June 2019 Overvoltage 12:30
5.3.2 Discussions (Test 2)
Consumer HC results in this test also varied monthly as observed in Test 1, however varied
inversely proportional to monthly solar irradiation levels. Results presented in Fig. 30 shows
that monthly HC results at a consumer-level ranged from 23% to 32%. These percentages
correspond to 3.17 kW and 4.4 kW, respectively. January and October had the lowest HC of
23%, while July had the highest HC of 32%. Results in Fig. 30 shows that HC at a network level
ranged from 67% to 71% of the substation transformer rating. Results in Table 12 show that
overvoltage was the main violated performance metric throughout the test. These overvoltage
conditions were encountered between 11:30 and 15:00. HC results obtained in this test shows
comparability to HC practices of countries reviewed in this study and HC results obtained in
similar studies for similar type of networks as summarized below:
a) Comparison of consumer HC results to practices in other countries and literature
- The minimum monthly HC of 23% of the standard utility service connection for
residential consumer compares to South Africa’s generation limit of 25% on
consumer’s notified maximum demand (NMD) for consumers in shared networks,
which are mostly residential consumers [85].
- The Maximum monthly HC of 32% on residential utility service connection (4.4 kW)
compares to Australia’s and Turkey’s maximum capacity of 5 kW (also used as an
export limit in Australia) recommended for DG systems in single-phase networks
[106], [107].
80
- The UK’s capacity of 3.68 kW (also used as export limit) recommended for DG
systems in single-phase networks [86], [87] compares to the obtained consumer HC
result range of 3.17 kW and 4.4 kW, (23% and 32% of 13.8 kW).
- The minimum monthly HC of 23% of residential utility service connection compares
to the residential consumer HC of 25% of consumer maximum load discovered in a
study conducted on a network in Virginia [93].
b) Comparison of network HC results to practices in other countries and literature
The network HC results ranging from 67% to 71% of the transformer nominal rating
compares to South Africa’s DG aggregate limit of 75% of MV/LV transformer rating [85],
and the acceptable aggregate DG capacity of 70% of installed transformer capacity in
typical urban European LV networks [18].
HC results in this test also varied monthly as observed in the previous test that considered
phase imbalance. Nonetheless, HC results from this test were higher than HC results obtained
in Test 1. This confirms that phase imbalance, which is an inherent characteristic of most single-
phase residential networks, is a limiting factor of the uptake of DGs in residential distribution
networks. HC results from this test were comparable to practices in other countries and results
in some literatures for similar type of networks, which might be confirming that considerations
of methods used to produce results applied in these practices are in the same ballpark as Test
2 in the context of disregarding phase unbalance. The differences in the HC results in Test 1
and Test 2 shows the importance of considering phase unbalance when conducting HC studies.
5.4 Testing HC methodology on a business distribution network (Test 3)
The HC assessment methodology was also tested on a business network, called Spar
distribution network, which supplies power to six consumers in a business area. At the time of
data collection, this network did not have any prosumer.
5.4.1 As-Built Drawing of Spar Network
Spar network is supplied from a secondary substation called Spar substation, which contains
a 630 kVA (11/0.42 kV) transformer. The distribution network supplies power to six consumers
81
through dedicated LV feeder cables. These consumers include ERF 4893, ERF 622, ERF 624,
ERF 2573, ERF 2020, and ERF 3105, as presented in Fig. 31.
Fig. 31: Cadastral layout of Spar network
Fig. 32: Network configuration of Spar network
82
Table 13: Properties of LV feeder cables in Spar Network
Label Type Length
(m)
Derating
factor
Cable 1 Cu 95mm2 4C PVC Cable 40 1
Cable 2 Cu 95mm2 4C PVC Cable 120 1
Cable 3 Cu 95mm2 4C PVC Cable 70 1
Cable 4 Cu 95mm2 4C PVC Cable 150 1
Cable 5 Cu 95mm2 4C PVC Cable 80 1
Cable 6 Cu 95mm2 4C PVC Cable 270 1
5.4.2 Model of Spar Network in DigSilent PowerFactory 2018
The Spar network (Fig. 32) was modeled in DigSilent PowerFactory 2018 according to the as-
built drawing received from Erongo RED. The model consisted of an MV/LV transformer and
dedicated LV feeder cables from the transformer to each consumer’s meter-room. Consumer
loads were defined using historical time-series metering data (with a time resolution of 30
minutes) covering a period of four months, from June 2019 to September 2019. The four-month
data was the only available and retrievable data from the meters due to the meters’ limited
storage capacity of 120 days. The geographical coordinates of !Nara network were defined in
DigSilent to provide the software access to meteorological data of this location.
Service connections for business consumers are generally defined in terms of notified
maximum demands (NMD) declared by consumers annually. According to the Namibian net
metering policy, a consumer is eligible to install a PV system with a capacity up to the
consumer’s service connection with the utility, defined by the consumer’s circuit breaker rating
or NMD. Each consumer in the model was therefore allocated a PV system with a nominal
capacity equivalent to the consumer’s NMD (Table 14). Inverters sized according to consumers’
NMDs were considered in the modelling of PV systems. The loads and PV systems were
modelled and defined as described in Chapter 4.
The DigSilent model of Spar network is shown in Fig. 33. Fig. 34 shows a typical outline of a
meter-room, illustrating how consumer loads and PV systems are connected in the model .
83
Table 14: NMD and PV capacity for consumers in Spar Network
Consumer NMD
(kVA)
Max PV Capacity
(kW)
ERF 4893 200 200
ERF 622 110 110
ERF 624 55 55
ERF 2020 55 55
ERF 2573 41 41
ERF 3105 103.8 103.8
Fig. 33: DigSilent Network Model for Spar Network
84
Fig. 34: Inside configuration of a meter-room model
5.4.3 Simulation Results and Discussion of the DG HC of Spar Network
Spar network model was simulated using the developed DPL script to produce network voltage
profiles and loading profiles of the transformer and LV feeder cables. In this test, simulations
did not generate script summary reports because no violation of performance metrics were
encountered from a GSF of 0% to 100%.
a) HC results for Spar Network
Voltage and loading profiles for Spar Network are provided in APPENDIX E and APPENDIX F.
APPENDIX E shows a typical LV voltage profile of a transformer and consumer meter-rooms
with reference to June 2019. APPENDIX F shows a typical loading profile of the transformer
and consumer feeder cables with reference to June 2019. Voltage and loading profiles for the
rest of the months have been submitted with this report as supplementary information.
The results shown in Table 15 were derived from the monthly loading profiles at the GSL =
100%. Column two of Table 15 presents the Spar network’s HC at a consumer level. The results
show that a maximum GSF of 100% was achieved without triggering technical constraints
(overvoltage and overload). The maximum GSF is considered as HC at a consumer level
relative to the consumer’s NMD.
Column three of Table 15 presents HC results at a network level. The maximum transformer
loading obtained at the maximum GSF during each month of the study period is considered as
HC at a network level. The maximum transformer loading during the four considered months
was below 60% of the transformer nominal rating, as shown in column three of Table 15.
Table 15: Hosting capacity results of Spar network
Month
Max GSF or Consumer DG hosting
capacity
(relative to consumer NMD)
(%)
Transformer loading
(at Max GSF)
(%)
June 2019 100% 39.35%
July 2019 100% 42.92%
85
August 2019 100% 50.37%
September 2019 100% 55.85%
b) Discussions of HC results for Spar Network
Spar network would not be severely impacted by DG integration sized according to the NMR.
The results confirms that consumer-side PV systems sized up to 100% of consumers’ notified
maximum demands (NMDs) could be achieved with a reduced risk of causing technical
constraints such as overvoltage and overload in a network that supplies power to consumers
connected via dedicated LV feeder cables, provided that the system is balanced. Business
prosumers supplied by dedicated feeders have moderate daytime demands, which allow them
to consume much of the energy generated by the PV systems with a reduced risk of grid
exports. The results would however be different if the substation had a large number of
consumers connected to it, which if the number was high would result in transformer overload
depending on the magnitude of grid exports.
The four months (June, July, August, and September) considered in this case study are
generally characterized by low irradiance levels. Since the consumer hosting capacity results
obtained in Test 2 (!Nara network) were inversely proportional to solar irradiance, we may
expect consumer HC results for the Spar network during high solar irradiance months to drop
slightly below the results for low solar irradiance months.
The results also show that distribution networks supplying consumers through dedicated
feeders connected directly from the substations are less susceptible to overvoltage constraints
when PV consumer-side DG are integrated into them.
Comparisons of the HC results to practices in other countries and literature are as follows:
i. Comparison of consumer HC results to practices in other countries and literature:
This study confirms that a generation capacity limit of 100% of the consumer’s NMD for
business consumers connected via balanced dedicated LV feeder cables is practical
without network constraints. These results slightly compare with South Africa’s
generation capacity limit of 75% of a consumer’s NMD for consumers connected to
dedicated networks [85]. The results also slightly compare with those published in a
86
study focused on assessing the impacts of PV DG using a real distribution network
located in Virginia; a hosting capacity of 55% of consumer maximum load was identified
as the optimum DG penetration level for a commercial consumer [93].
ii. Comparison of network HC results to practices in other countries and works of other
literature:
Network HC results obtained from this study were below 60% of the transformer nominal
rating, which shows that there is still more room for DG integration. Practices in South
Africa and Europe limit aggregate DG capacity to 75% and 70% of the transformer
nominal rating, respectively [18], [85].
5.5 Applications of HC Results
a) The monthly consumer HC results (minimum or maximum) from this methodology may
be used by DNOs to either restrict sizing of DG systems or as a guideline to impose
grid export limits for DG systems sized according to the current regulatory policy (NMR).
The latter is more favorable because it gives prosumers the flexibility to optimize their
electricity use by matching their generation to their demands and manage their excess
generations through either energy storage or power curtailment as per export limits.
DNOs can impose export limits during commissioning of DG systems.
b) The maximum monthly transformer HC result obtained from this methodology may be
used as an aggregate generation capacity limit for DG systems approved for grid export
and net metering. Once this capacity is reached, then new DGs may be approved as a
non-exporting DGs. The reduced grid energy exports will minimize the risk of adverse
technical impacts in distribution networks.
c) Results obtained in this study calls for two innovation opportunities for inverter
manufacturers. The two innovation opportunities have the potential help most countries
abolish restrictive DG sizing measures, which limits the uptake of DGs in distribution
networks, to allow prosumers to have optimally sized DGs to meet their energy
requirements while keeping the networks safe. These two innovation opportunities are
described below:
87
i) Seeing that technical constraints are encountered frequently during hours of
high solar irradiation, it would add value if inverter manufacturers can develop
inverters that automatically switches into ‘load following mode’ (produce
according to on-site demand) or automatically block exports to the grid during
hours of high solar irradiation to minimize technical constraints associated with
grid exports. This will give prosumers the flexibility to optimize and match their
generation to their demands. Or,
ii) Seeing that HC results vary monthly, it would also add value if inverter
manufacturers can develop inverters with a date-based export limit functionality.
This function will allow utilities to impose monthly export limits according to
monthly HC limits obtained from technical studies to minimize technical
constraints, while giving prosumers the flexibility to optimize and match their
generation to their demands. DNOs can impose these limits during
commissioning of DG systems.
5.6 The relevance of Developed HC Methodology
The script-based methodology for assessing hosting capacities (HC) in distribution networks
was successfully developed and tested on two types of distribution networks (residential and
business). The methodology produced varying monthly HC results as anticipated due to month-
to-month variation in demands and solar irradiation. A residential network is generally expected
to have low HC as compared to a business network and this has been confirmed by the results.
5.7 Limitation of HC Methodology and Results
a) The application of this methodology is limited to existing distribution networks with
available historical time-series load data for more than a month, where a high and
uniform uptake of DG is expected. Probabilistic methods are recommended for
applications with high degree of uncertainty in load/generation profiles, as well as,
uncertainty on capacity and location of DGs, especially when designing and planning
new distribution networks which anticipate high uptake of DG.
88
b) The accuracy of the HC results depends on the availability of load data and the accuracy
of the network model.
c) The methodology requires DigSilent modelling skills and programming knowledge in
either C, C++, or python programming languages.
d) Depending on the volume of the metering data, the methodology can be very slow to
execute (a day and a half for a 12 month data) on a general-purpose computer. The
speed can be enhanced by using a high-performance computer.
e) Acknowledging the fact that to date there is no inverter featured with the date-based
multiple export limit functionality to accommodate the application of monthly HC through
monthly grid export limits; this study introduces an innovation opportunity for inverter
manufacturers to consider in their future inverter innovations.
89
6. CONCLUSION
A post-implementation review of the Namibian NMR was conducted. The review indicated that
consumer-side DG export large volumes of energy to distribution networks between 12:00 -
14:30 during non-winter months. These exports bring financial benefits to both prosumers, and
DNOs when compensated at the avoided cost. This changed the narrative of considering NM
compensations as a loss to DNOs, except when grid exports are credited at the ratio of 1:1.
Grid exports from prosumers in Namibia is therefore a cheaper source of energy to DNOs as
compared to the national utility that charges other energy service charges i.e. reliability charge,
transmission losses charge etc. on top of the ‘avoided cost’. DNOs are however expected to
lose revenues not because of NM compensations, but because of reduced volumes of energy
sales caused by the reduction in energy consumption from the grid when prosumers offset their
demand with generation in real-time. Additional revenue loss is expected as a result of peak
shaving among business prosumers, whose fixed charges are linked to demand peaks. Large
volumes of grid exports were also shown to increase the risk of technical constraints in
distribution networks. Adaptations to current utility and regulatory policies were proposed to
address the financial and technical risks posed on DNOs by consumer-side DG.
A methodology for assessing the HC of DG in LV distribution networks was developed and
tested in a residential and business distribution network. Results from the tests show that the
adequacy of networks to host DGs is network specific, and depends on several factors,
including consumer load profiles, sizes and location of DGs, concentration of DGs, network
characteristics, climatic conditions, conductor sizes, phase balancing etc. The results further
demonstrated the importance of considering phase unbalance, which causes voltage and
loading unbalance, especially when conducting HC studies for residential distribution networks.
The developed methodology could help DNOs facing high uptake of DGs in existing distribution
networks in Namibia and other countries in Africa evaluate the adequacies of their existing
distribution networks to host new DGs, while acknowledging the presence of existing
prosumers. This methodology is however limited to existing distribution networks expecting
high and uniform uptake of DGs, especially in business areas and middle/high income
residential areas. In the absence of historical load data, DNOs should consider probabilistic
methods when evaluating HC in new distribution networks or existing distribution networks
90
characterized with uncertainties in load and generation profiles, as well as, location and
capacities of DGs.
Recommendations are further provided to inform DNOs, regulators, and inverter manufacturers
of necessary adaptions that needs consideration to allow the ESI to embrace consumer-side
DGs without compromising the reliability of distribution networks. The proposed adaptions
acknowledge the risk of funds inadequacy faced by DNOs to maintain electrical infrastructures
and mitigate technical constraints caused by DGs.
The research questions, which were the fundamental drivers of this study and the yardsticks
for measuring its success, are answered as follows:
Question 1: What is the status quo of Net Metering in Namibia?
Answer:
The net metering policy as an incentive for consumer-side DG has significantly boosted
consumer preference for self-generation in Namibia. According to the information obtained
from DNOs, Namibia’s current developments for consumer-side DG stands at 881
installations with a combined installed capacity of 40.94 MW, which is 6.3% of Namibia’s
highest maximum demand of 653 MW. These DG export large volumes of energy to
distribution networks. Through net metering, these exports bring significant financial
benefits to prosumers. The implementation aftermaths of the policy however left DNOs
skeptical about whether the policy is suited for long-term application. The first concern is
the policy’s recommended generation capacity limit (based on a consumer’s service circuit
breaker rating or NMD) which poses a high risk of technical constraints. The second
concern is the policy’s recommended capacity dissemination methodology based on a
“first-come-first-serve” basis, which lacks consumer inclusivity. To address these
shortcomings, cost-effective mechanisms (alternative to the existing) are required to help
DNOs distribute and allocate DG hosting capacities to prosumers in a fair, inclusive, and
technically sustainable manner. This will ensure a safe and reliable operation of distribution
networks upon DG integration, without requiring costly network reinforcements. This
acknowledges the risk of funds inadequacy faced by DNOs to maintain electrical
infrastructures and mitigate technical constraints caused by DGs.
91
Question 2: What are the impacts of consumer-side DGs in distribution networks?
Answer:
Consumer-side DG export large volumes of energy to distribution networks. These exports
cause technical constraints such as overvoltage, voltage unbalance and overload
conditions due to network capacity bottlenecks caused by the design philosophy of
distribution networks. These constraints may affect the reliability of distribution networks
and the quality of supply to end-users, requiring costly mitigation measures. Increasing
number of consumer-side DG will cause DNOs to lose considerable revenues due to
reduced volumes of energy sold to prosumers when they offset their demands with their
generation in real-time to reduce consumption from the grid.
Question 3: What adaptations to current utility and regulatory policies are required to
encourage a sustainable deployment of consumer-side DGs?
Answer:
The following adaptations were recommended to encourage a sustainable deployment of
consumer-side DGs:
i) To address the issue of consumer inclusivity, the DG capacity dissemination
methodology recommended by the policy needs to be amended in such a way that
it considers all consumers in areas with high potential for DG uptake as prospective
prosumers to give them an equal opportunity to become prosumers. To
accommodate this provision, DNOs must ensure that individual generation capacity
limits or grid export limits and networks’ aggregate generation limits are set within
the hosting capacities of distribution networks. This will ensure that the move to
allow all consumers to become prosumers, which is set to happen inevitably, will
not compromise the reliability of distribution networks or bring costly network
reinforcement implications.
ii) The policy should encourage DNOs to conduct technical studies of their distribution
networks to determine the maximum amounts of DG that could be integrated
without causing adverse technical impacts.
92
iii) DNOs should encourage consumers to install inverters with ‘export limit functions’
for ease of imposing grid export limit requirements according to hosting capacity
results.
iv) DNO’s should consider redesigning their traditional tariff structures to acknowledge
the presence of DG and eliminate cross-subsidization of prosumers by normal
consumers. This can be achieved by certainly reallocating a portion of the fixed
cost recovery from the volumetric kWh charge to a fixed charge.
v) The net metering policy needs to acknowledge the risk of funds inadequacy faced
by DNOs to maintain electrical infrastructures and mitigate technical constraints
caused by DGs.
vi) DNOs should design or acquire computerized DG databases to track the number
and capacity of grid connected DGs at various interconnection points for reporting
purposes and decision-makings in the approval process of DGs and application of
HC results.
These adaptions are further discussed in the next chapter.
Question 4: What methodology can DNOs use to assess the DG hosting capacity of their
distribution networks?
Answer:
A script-based methodology for assessing hosting capacity in LV distribution networks was
developed and tested in this dissertation. The methodology was developed using DigSilent
Programming Language (DPL), a tool in DigSilent PowerFactory. DNOs may use this
methodology to calculate hosting capacity results for their distribution networks. The pros and
cons of this methodology are defined in this dissertation. In the absence of historical load data,
DNOs should consider probabilistic methods when evaluating HC in new distribution networks
or existing distribution networks characterized with uncertainties in load and generation profiles,
as well as, location and capacities of DGs.
93
Question 5: How can DNOs utilize DG hosting capacity results to implement operational
strategies to ensure a sustainable integration of consumer-side DGs?
Answer:
1) The concept of hosting capacity will encourage DNOs to standardize the use of
inverters with ‘export limit functions’ to help consumers respond to temporal or
permanent export limit requirements imposed by the DNO.
2) DNOs may use hosting capacity results to create interconnection guidelines defining
DG sizing requirements, export limit requirements, and aggregate generation capacity
requirements.
i) In the case of Namibia, DNOs may use consumer HC results calculated on a
monthly basis to define monthly grid export limits (curtailment requirements) for
new DGs that will be sized according to the current net metering policy.
The practical implementation of monthly HC require upgrades to existing
inverter technology, which currently contains a single export limit functionality.
This opens the possibility to drive innovation in the inverter technology, to
develop a date-based multiple export limit functionality.
ii) DNOs may restrict the sizing of new DG systems to the minimum monthly
consumer HC for consumers intending to use inverters without ‘export limit
functions’.
iii) Current inverters with an export limit functionality allows a single fixed export
limit to be defined. DNOs may therefore impose grid export limits according to
the minimum monthly consumer HC for new DGs that will be sized according to
the current net metering policy.
iv) DNOs may use the maximum network hosting capacity result to define an
aggregate capacity limit for DG systems approved for grid exports and net
metering. Once this capacity is reached, DNOs can resort to approving all new
DG systems as non-exporting and non-net metering DGs.
94
7. RECOMMENDATIONS
This study recommends the following:
a) DNOs should conduct technical studies to assess the hosting capacities of DG in their
respective distribution networks. HC results will help them develop effective strategies
to accommodate more prosumers without compromising the reliability of distribution
networks.
b) DNOs should use HC results to either restrict DG sizing or impose grid export limits,
especially in residential distribution networks. This will minimize the risk of causing
technical constraints and ensure a safe and reliable operation of distribution networks
upon DG integration. Violations of export restrictions may be penalized with a graduated
charge scheme for power exports exceeding a set threshold.
c) DNOs should standardize the use of inverters with ‘export limit functions’ to help
consumers respond to temporal or permanent export limit requirements imposed by the
DNO.
d) Inverters with the ‘export limit functions’ allows a single fixed export limit to be defined.
DNOs may therefore impose grid export limits according to the minimum monthly
consumer HC for new DGs sized according to the current net metering policy.
e) To date there is no inverter with the date-based export limit functionality. Inverter
manufacturers should assess the feasibility and practicality to develop inverters with a
date-based export limit functionality. This function will allow DNOs to impose monthly
export limits according to monthly HC limits obtained from technical studies to minimize
technical constraints, while giving prosumers the flexibility to optimize and match their
generation to their demands.
f) Inverter manufacturers should also assess the feasibility and practicality to develop
inverters, which automatically switches into ‘load following mode’ (produce according to
95
on-site demand) or automatically block exports to the grid during hours of high solar
irradiation to minimize technical constraints associated with grid exports.
g) DNOs must only offer net metering credits to consumers connected to distribution
networks that physically allow DNOs to redistribute or resell energy exports to other
consumers to prevent revenue losses.
h) To address the issue of cross-subsidization brought upon by self-generation and net
metering policies, DNOs and regulators must change their electricity rate design by
certainly reallocating a portion of the fixed cost recovery from the volumetric kWh charge
to a fixed charge or net metering charge.
i) DNOs and regulators that have implemented net metering policies must conduct policy
post-implementation reviews to assess their effectiveness and incorporate lessons
learned.
j) The electricity regulator in Namibia, who is the custodian of the net metering policy,
needs to acknowledge the risk of funds inadequacy faced by DNOs to maintain
electrical infrastructures and mitigate technical constraints caused by DGs. This must
be reflected in the policy by defining measures that will minimize the export of energy
from DGs to the grid, either through strict generation capacity limits or through export
limits to ensure that DGs are within network hosting capacities.
k) DNOs should design or acquire computerized DG databases to track the number and
capacity of grid connected DGs at various interconnection points for reporting purposes
and decision-makings in the approval process of DGs and application of HC results.
l) DNOs should consider probabilistic methods when evaluating HC in new distribution
networks or existing distribution networks characterised by uncertainties in load and
generation profiles, as well as, location and capacities of DGs.
m) Other African countries in the process of implementing Net Metering policies should
unbundle their generation and transmission tariffs, and consider ‘avoided cost’ as
means to compensate grid exports from consumer-side DGs. This will reduce the risk
of DNOs losing revenues as a result of net metering compensations.
96
n) DNOs should not resort to increasing electricity tariff rates in the hope to compensate
revenue losses incurred as a result of self-generation because high tariffs don’t solve
the problem but rather make net metering policies more attractive and spur further
deployment of consumer-side DG, which eventually exacerbate the problem. DNOs
should rather find cheaper alternative source of energy to help them reduce or discount
their tariffs to end-users without compromising their profits. Securing affordable supply
rates either through PPAs or through own funded generation projects can achieve this.
97
REFERENCES
[1] T. R. Ricciardi, K. Petrou, J. F. Franco, and L. F. Ochoa, “Defining Customer Export
Limits in PV-Rich Low Voltage Networks,” IEEE Trans. Power Syst., vol. 34, no. 1, pp.
87–97, Jan. 2019, doi: 10.1109/TPWRS.2018.2853740.
[2] A. Meddeb, N. Sahbeni, H. Jmii, and S. Chebbi, “Impact of distributed generation on
the protection system in Tunisian distribution network,” in 2018 15th International Multi-
Conference on Systems, Signals and Devices, SSD 2018, 2018, pp. 514–520, doi:
10.1109/SSD.2018.8570584.
[3] Z. Wang, J. Li, W. Yang, and Z. Shi, “Impact of Distributed Generation on the power
supply reliability,” in IEEE PES Innovative Smart Grid Technologies, 2012, pp. 1–5,
doi: 10.1109/ISGT-Asia.2012.6303172.
[4] P. Chiradeja, “Benefit of Distributed Generation: A Line Loss Reduction Analysis,” in
2005 IEEE/PES Transmission & Distribution Conference & Exposition: Asia and
Pacific, 2005, vol. 2005, pp. 1–5, doi: 10.1109/TDC.2005.1546964.
[5] A. Amiri et al., “Thermal aspects of a low cost power electronic converter for high
capacity, smart residential distribution networks,” in 8th IET International Conference
on Power Electronics, Machines and Drives (PEMD 2016), 2016, pp. 1–6, doi:
10.1049/cp.2016.0281.
[6] E. T. Fasina, A. S. Hassan, and L. M. Cipcigan, “Impact of Localised Energy
Resources on electric power distribution systems,” in 2015 50th International
Universities Power Engineering Conference (UPEC), 2015, pp. 1–5, doi:
10.1109/UPEC.2015.7339793.
[7] X. Jia, H. Du, H. Zou, and G. He, “Assessing the effectiveness of China’s net-metering
subsidies for household distributed photovoltaic systems,” J. Clean. Prod., vol. 262, no.
20 July 2020, pp. 1–9, Jul. 2020, doi: 10.1016/j.jclepro.2020.121161.
[8] Y. Ji, M. Hallack, A. Novaes Mejdalani, J. E. Chueca, and D. D. Lopez Soto,
“Implementing net metering policies in latin america and the caribbean: design,
incentives and best practices,” Washington, D.C., Dec. 2018. doi: 10.18235/0001463.
98
[9] B. Banfield, P. Ciufo, and D. A. Robinson, “The technical and economic benefits of
utility sponsored renewable energy integration,” in 2017 Australasian Universities
Power Engineering Conference (AUPEC), 2018, pp. 1–6, doi:
10.1109/AUPEC.2017.8282489.
[10] R. Dufo-López and J. L. Bernal-Agustín, “A comparative assessment of net metering
and net billing policies. Study cases for Spain,” Energy, vol. 84, pp. 684–694, May
2015, doi: 10.1016/j.energy.2015.03.031.
[11] P. Celvakumaran, V. K. Ramachandaramurthy, and J. Ekanayake, “Assessment of Net
Energy Metering on Distribution Network Losses,” in 2019 IEEE International
Conference on Automatic Control and Intelligent Systems (I2CACIS), 2019, pp. 1–6,
doi: 10.1109/I2CACIS.2019.8825071.
[12] W. Geffert and K. Strunk, “Beyond net metering: A model for pricing services provided
by and to distributed generation owners,” Electr. J., vol. 30, no. 3, pp. 36–43, Apr.
2017, doi: 10.1016/j.tej.2017.02.007.
[13] J. Ossenbrink, “How feed-in remuneration design shapes residential PV prosumer
paradigms,” Energy Policy, vol. 108, pp. 239–255, 2017, doi:
10.1016/j.enpol.2017.05.030.
[14] N. R. Darghouth, R. H. Wiser, G. Barbose, and A. D. Mills, “Net metering and market
feedback loops: Exploring the impact of retail rate design on distributed PV
deployment,” Appl. Energy, vol. 162, 2016, doi: 10.1016/j.apenergy.2015.10.120.
[15] Solar Electric Power Association, “Ratemaking, solar Value and solar net energy
metering - a primer,” Washington, 2013. [Online]. Available:
https://sepapower.org/resource/ratemaking-solar-value-and-net-energy-metering-
primer/.
[16] Y. Parag and B. K. Sovacool, “Electricity market design for the prosumer era,” Nat.
Energy, vol. 1, no. 4, p. 16032, Apr. 2016, doi: 10.1038/nenergy.2016.32.
[17] M. S. S. Abad, G. Verbic, A. Chapman, and J. Ma, “A linear method for determining the
hosting capacity of radial distribution systems,” in 2017 Australasian Universities
Power Engineering Conference (AUPEC), 2017, vol. November 2, pp. 1–6, doi:
10.1109/AUPEC.2017.8282428.
[18] T. Aziz and N. Ketjoy, “PV Penetration Limits in Low Voltage Networks and Voltage
99
Variations,” IEEE Access, vol. 5, pp. 16784–16792, 2017, doi:
10.1109/ACCESS.2017.2747086.
[19] Ministry of Mines and Energy (MME), “National renewable energy policy (NREP),”
Windhoek, 2017. [Online]. Available:
http://www.mme.gov.na/files/publications/03f_National Renewable Energy Policy - July
2017.pdf.
[20] E. Mulenga, M. H. J. Bollen, and N. Etherden, “A review of hosting capacity
quantification methods for photovoltaics in low-voltage distribution grids,” Int. J. Electr.
Power Energy Syst., vol. 115, no. 2020, pp. 1–13, 2020, doi:
10.1016/j.ijepes.2019.105445.
[21] A. M. Simeon and T. Wanjekeche, “The Impacts of Increased Penetration of
Distributed Generators on the Namibian Grid Stability,” 2018, doi:
10.1109/OI.2018.8535966.
[22] A. M. Simeon, T. Wanjekeche, E. Hamatwi, and N. I. Kamati Mbangula, “Impacts of
Distributed Energy Resources on the Grid Transient Stability: A Case Study of
Namibia,” 2019, doi: 10.1109/OI.2019.8908179.
[23] A. M. Simeon, T. Wanjekeche, and E. Hamatwi, “Impacts of increased integration of
Wind and Solar generators on the Namibian grid power losses,” 2019, doi:
10.1109/PowerAfrica.2019.8928866.
[24] National planning Commission, “Namibia’s 5th National Development Plan (NDP 5),”
Windhoek, 2017. [Online]. Available:
http://www.gov.na/documents/10181/14226/NDP+5/.
[25] NamPower, “NamPower annual report 2016,” Windhoek, 2016. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/NamPower AR
09.03.20_web.pdf.
[26] NamPower, “NamPower annual report 2017,” Windhoek, 2017. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/Nampower Annual Report
2017 FA Approved LR-2.pdf.
[27] Ministry of Mines and Energy (MME), “National energy policy (NEP),” Windhoek, 2017.
[Online]. Available: http://www.mme.gov.na/files/publications/fd8_National Energy
Policy - July 2017.pdf.
100
[28] Ministry of Mines and Energy (MME), “National integrated resource plan (NIRP) review
and update,” Windhoek, 2016. [Online]. Available:
http://www.mme.gov.na/files/publications/f74_29a_ELECTRICITY SECTOR
NATIONAL INTEGRATED RESOURCE PLAN (NIRP) 2016 FINAL REPORT.pdf.
[29] Electricity Control Board (ECB), “Electricity Control Board (ECB) annual report 2017,”
Windhoek, 2017. [Online]. Available: https://www.ecb.org.na/images/docs/Annual
Reports/ECB_Annual_Report_2017.pdf.
[30] Electricity Control Board (ECB), “Electricity Control Board (ECB) annual report 2019,”
Windhoek, 2019. [Online]. Available: https://www.ecb.org.na/images/docs/Annual
Reports/ECB_Annual_Report_2019.pdf.
[31] NamPower, “NamPower annual report 2018,” Windhoek, 2018. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/Nampower 2018 Annual
Report 28.01.19-print.pdf.
[32] NamPower, “NamPower Annual Report 2010,” Windhoek, 2010. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-
reports/NomPower_Annual_Report_2010.pdf.
[33] NamPower, “NamPower annual report 2011,” Windhoek, 2011. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/Final Annual Report.pdf.
[34] NamPower, “NamPower annual report 2012,” Windhoek, 2012. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/PNP013 Front.pdf.
[35] NamPower, “NamPower annual report 2013,” Windhoek, 2013. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/NP_AR2013_Final.pdf.
[36] NamPower, “NamPower annual report 2014,” Windhoek, 2014. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/Nampower Annual Report
2014.pdf.
[37] NamPower, “NamPower annual report 2015,” Windhoek, 2015. [Online]. Available:
www.nampower.com.na.
[38] NamPower, “NamPower annual report 2019,” Windhoek, 2019. [Online]. Available:
https://www.nampower.com.na/public/docs/annual-reports/NamPower AR
09.03.20_web.pdf.
[39] Electricity Control Board (ECB), “Net metering rules: electricity act, 2007,” Windhoek,
101
2016. [Online]. Available: https://www.ecb.org.na/images/docs/Noticeboard/Gazetted
Net metering Rules.pdf.
[40] C. H. Ng, T. T. Lie, and L. Goel, “Impacts of distributed Generation on system reliability
in competitive electricity markets,” in 8th International Power Engineering Conference,
IPEC 2007, 2007, pp. 735–740, [Online]. Available:
https://ieeexplore.ieee.org/document/4510123.
[41] E. Stewart, J. MacPherson, S. Vasilic, D. Nakafuji, and T. Aukai, “Analysis of high-
penetration levels of photovoltaics into the distribution grid on oahu, hawaii: detailed
analysis of heco feeder wf1,” Colorado, 2013. [Online]. Available:
http://www.osti.gov/bridge.
[42] Chowdhury, Crossley, and Chowdhury, Microgrids and active distribution networks, 1st
ed. London: Institution of Engineering and Technology, 2009.
[43] D. W. Gao, Energy storage for sustainable microgrid. Waltham: Elsevier, 2015.
[44] Eskom, “NRS 034 - Electricity distribution : guidelines for the provision of electricity
distribution networks in residential areas. Part 1, Planning and design of distribution
networks.,” Technology Standardization Department (TSD), Johannesburg, 2007.
[45] C. Bayliss and B. Hardy, Transmission and distribution electrical engineering, 4th ed.
Waltham, MA 02451, USA: Elsevier Ltd, 2012.
[46] Lakervi and Holmes, Electricity Distribution Network Design, 2nd ed. London: The
Institution of Engineering and Technology (IET), 2003.
[47] M. Sun, I. Konstantelos, and G. Strbac, “Analysis of diversified residential demand in
London using smart meter and demographic data,” in IEEE Power and Energy Society
General Meeting, Nov. 2016, vol. 2016-Novem, doi: 10.1109/PESGM.2016.7741076.
[48] Eskom, “NRS 048-2: Electricity supply - quality of supply part 2 : voltage characteristics
, compatibility levels , limits and assessment methods,” Technology Standardization
Department (TSD), Johannesburg, 2003. [Online]. Available: www.stansa.co.za.
[49] S. Singh, J. Østergaard, and N. Jain, “Distributed generation in power systems: An
overview and key issues,” 24rth Indian Eng. Congr. NIT Surathkal, Kerala, December
10-13, 2009, pp. 1–8, 2009, [Online]. Available:
https://www.semanticscholar.org/paper/Distributed-Generation-in-Power-Systems-
%3A-An-and-Østergaard-Jain/2bf2ee130c950d6c4ad6ce8fd5388e0a5c08d54d.
102
[50] M. Geidl, “Protection of power systems with distributed generation: state of the art,”
Zurich, 2005. [Online]. Available:
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.464.7598&rep=rep1&type=
pdf.
[51] T. Ackermann, G. Andersson, and L. Söder, “Distributed generation: a definition,”
Electr. Power Syst. Res., vol. 57, no. 3, pp. 195–204, 2001, doi: 10.1016/S0378-
7796(01)00101-8.
[52] U. Shahzad, S. Kahrobaee, and S. Asgarpoor, “Protection of distributed generation:
challenges and solutions,” Energy Power Eng., vol. 09, no. 10, pp. 614–653, 2017, doi:
10.4236/epe.2017.910042.
[53] Energy Networks Association, “Distributed generation connection guide: A guide for
connecting generation to the distribution network in multiple premises that falls under
g83/2,” London, 2014. [Online]. Available: www.nie.co.uk.
[54] W. El-Khattam and M. M. A. Salama, “Distributed generation technologies, definitions
and benefits,” Electr. Power Syst. Res., vol. 71, no. 2, pp. 119–128, Oct. 2004, doi:
10.1016/j.epsr.2004.01.006.
[55] R. Viral and D. K. Khatod, “Optimal planning of distributed generation systems in
distribution system: A review,” Renew. Sustain. Energy Rev., vol. 16, no. 7, pp. 5146–
5165, Sep. 2012, doi: 10.1016/j.rser.2012.05.020.
[56] R. Seguin, J. Woyak, D. Costyk, J. Hambrick, and B. Mather, “High-penetration PV
integration handbook for distribution engineers,” Denver West Parkway, Golden, 2016.
doi: 10.2172/1235905.
[57] S. Kumar and R. Kumar, “Review of the distributed generation concept using PSO:
attempt of unification,” Int. J. Res. Rev. Eng. Sci. Technol., vol. 6, no. 1, p. 5, 2017,
[Online]. Available: https://doi.org/10.24084/repqj03.275.
[58] J. Widén, “Distributed photovoltaics in the swedish energy system,” Uppsala
University, 2009.
[59] B. Zhao, C. Wang, and X. Zhang, Grid-Integrated and Standalone Photovoltaic
Distributed Generation Systems - Analysis, Design, and Control. 111 River Street,
Hoboken, NJ07030, USA: John Wiley & Son, 2017.
[60] S. Yamujala, M. Fatima, I. A. S. Teja, and Y. Lahari, “Present scenario of distributed
103
generation in India - Technologies, cost analysis & power quality issues,” in
Proceedings of the International Conference on Innovative Applications of
Computational Intelligence on Power, Energy and Controls with Their Impact on
Humanity, CIPECH 2014, 2014, pp. 417–421, doi: 10.1109/CIPECH.2014.7019089.
[61] L. I. Duləu, M. Abrudean, and D. Bicə, “Impact of distributed generation upon Reghin -
Ləpuşna Medium Voltage line,” in 2014 International Symposium on Fundamentals of
Electrical Engineering (ISFEE), 2015, pp. 1–5, doi: 10.1109/ISFEE.2014.7050601.
[62] T. A. Short, Electric Power Distribution Handbook, 2nd ed. Boca Raton, Florida: CRC
Press LLC, 2018.
[63] H. Musa, “A Review of Distributed Generation Resource Types and Their Mathematical
Models for Power Flow Analysis,” Int. J. Sci. Technol. Soc., vol. 3, no. 4, 2015, doi:
10.11648/j.ijsts.20150304.21.
[64] M. C. Vargas, M. A. Mendes, and O. E. Batista, “Impacts of high PV penetration on
voltage profile of distribution feeders under brazilian electricity regulation,” in 2018 13th
IEEE International Conference on Industry Applications, INDUSCON 2018 -
Proceedings, 2019, pp. 38–44, doi: 10.1109/INDUSCON.2018.8627070.
[65] M. G. Villalva, J. R. Gazoli, and E. R. Filho, “Comprehensive approach to modeling and
simulation of photovoltaic arrays,” IEEE Trans. Power Electron., vol. 24, no. 5, pp.
1198–1208, 2009, doi: 10.1109/TPEL.2009.2013862.
[66] A. Luque and S. Hegedus, Handbook of Photovoltaic Science and Engineering. 2011.
[67] Electric Schneider, Electrical installation guide According to IEC international
standards Electrical installation guide, 2016th ed. Electric Schneider, 2016.
[68] International Renewable Energy Agency (IRENA), “Renewable Capacity 2019,” 2019.
[Online]. Available: www.irena.org/publications.
[69] N. R. Babu and P. Arulmozhivarman, “Wind energy conversion systems - A technical
review,” J. Eng. Sci. Technol., vol. 8, no. 4, 2013.
[70] H. S. Kim and D. D.-C. Lu, “Wind Energy Conversion System from Electrical
Perspective—A Survey,” Smart Grid Renew. Energy, vol. 01, no. 03, 2010, doi:
10.4236/sgre.2010.13017.
[71] G. T. Tucho, “Feasible Biomass Energy Conversion Technologies In Developing
Countries,” Int. J. Eng. Res. Technol., vol. 2, no. 6, pp. 2720–2728, 2013, doi: 2278-
104
0181.
[72] N. S. Gopiya, D. K. Khatod, and M. P. Sharma, “Distributed Generation Impact on
Distribution Networks : a Review,” Int. J. Electr. Electron. Eng., vol. 2, no. 1, pp. 68–72,
2012.
[73] A. Faaij, “Modern biomass conversion technologies,” in Mitigation and Adaptation
Strategies for Global Change, 2006, vol. 11, no. 2, doi: 10.1007/s11027-005-9004-7.
[74] M. Siedlecki, W. de Jong, and A. H. M. Verkooijen, “Fluidized bed gasification as a
mature and reliable technology for the production of bio-syngas and applied in the
production of liquid transportation fuels-a review,” Energies, vol. 4, no. 3. 2011, doi:
10.3390/en4030389.
[75] M. Mes, G. M. A. Vanalme, J. M. A. Myrzik, M. Bongaerts, G. J. P. Verbong, and W. L.
Kling, “Distributed generation in the Dutch LV network self-supporting residential area,”
in 2008 43rd International Universities Power Engineering Conference, 2008, p. 5, doi:
10.1109/UPEC.2008.4651555.
[76] M. A. Mahmud, M. J. Hossain, and H. R. Pota, “Analysis of voltage rise effect on
distribution network with distributed generation,” in IFAC Proceedings Volumes (IFAC-
PapersOnline), 2011, vol. 44, no. 1 PART 1, doi: 10.3182/20110828-6-IT-1002.01305.
[77] P. Celvakumaran, V. K. Ramachandaramurthy, S. Padmanaban, K. Padmanathan, A.
Pouryekta, and J. Pasupuleti, “Technical constraints of integrating net energy metering
from the Malaysian perspective,” in Asia-Pacific Power and Energy Engineering
Conference, APPEEC, 2018, vol. 2018-October, doi: 10.1109/APPEEC.2018.8566496.
[78] International Electrotechnical Commission, “IEC 61727:Photovoltaic (PV) systems –
characteristics of the utility interface,” Order A J. Theory Ordered Sets Its Appl., vol.
2002, 2004.
[79] IEEE Standards Coordinating Committee 21, “IEEE standard for interconnecting
distributed resources with electric power systems,” Institute of Electrical and
Electronics Engineers (IEEE), New York, 2003. [Online]. Available:
https://standards.ieee.org/products-services/standards-related/pdf/electric-power-
systems.html.
[80] National Renewable Energy Laboratory(NREL), “DG power quality , protection and
reliability case studies report,” Newyork, 2003. [Online]. Available:
105
http://www.osti.gov/bridge.
[81] A. Picciariello, K. Alvehag, and L. Soder, “Impact of network regulation on the incentive
for DG integration for the DSO: opportunities for a transition toward a smart grid,” IEEE
Trans. Smart Grid, vol. 6, no. 4, pp. 1730–1739, Jul. 2015, doi:
10.1109/TSG.2015.2409313.
[82] M. Castaneda, M. Jimenez, S. Zapata, C. J. Franco, and I. Dyner, “Myths and facts of
the utility death spiral,” Energy Policy, 2017, doi: 10.1016/j.enpol.2017.07.063.
[83] M. S. S. T. K. Nagsarkar, Power System Analysis, 2nd ed. Oxford University Press,
2014.
[84] Eskom, “NRS 097-02-01 - Grid interconnection of embedded generation. Part 2, small-
scale embedded generation. section 1, utility interface.,” The Standardization Section,
Johannesburg, 2010.
[85] Eskom, “NRS 097-02-03 - Grid interconnection of embedded generation. Part 2, small-
scale embedded generation. section 3, simplified utility connection criteria for low-
voltage connected generators,” The Standardization Section, Johannesburg, 2010.
[Online]. Available: http://www.sabs.co.za.
[86] Energy Networks Association (ENA), “Engineering recommendation G98: requirements
for the connection of fully type tested micro-generators (up to and including 16 A per
phase) in parallel with public low voltage distribution networks,” London, 2018. [Online].
Available: https://www.energynetworks.org/electricity/engineering/distributed-
generation/engineering-recommendation-g98.html.
[87] Energy Networks Association (ENA) and Energy Networks Association, “Engineering
recommendation G99 - requirements for the connection of generation equipment in
parallel with public distribution networks on or after 27 April 2019,” no. 1, p. 67, 2019,
[Online]. Available: https://www.energynetworks.org/electricity/engineering/distributed-
generation/engineering-recommendation-g99.html.
[88] Energy Networks Association (ENA), “Engineering recommendation G100 Issue 1
2016 - technical guidance for customer export limiting schemes,” London, 2016.
[Online]. Available: www.energynetworks.org.
[89] Energy Networks Australia (ENA), National distributed energy resources grid
connection guidelines technical guidelines for low voltage EG connections. Melbourne
106
VIC 3000, 2018.
[90] Energy Networks Australia (ENA), National distributed energy resources grid
connection guidelines technical guidelines for basic micro EG connections. Melbourne
VIC 3000, 2019.
[91] Distributed Utility Associates and Endecon Engineering, “Distributed generation
interconnection manual public utility commission of Texas,” Livermore, CA 94550,
2001. [Online]. Available: www.endecon.com.
[92] A. R. Ahmadi, T. Manandhar, J. Barros, M. Bernardo, and S. Georgiopoulos, “UK
power networks’ experience of managing flexible distributed generation from planning
to operation,” in 24th International Conference & Exhibition on Electricity Distribution
(CIRED), 2017, vol. 2017, no. 1, pp. 2032–2036, doi: 10.1049/oap-cired.2017.1085.
[93] M. H. Athari, Z. Wang, and S. H. Eylas, “Time-series analysis of photovoltaic
distributed generation impacts on a local distributed network,” in 12th IEEE PES
PowerTech Conference, 2017, pp. 1–6, doi: 10.1109/PTC.2017.7980908.
[94] L. K. L. Estorque and M. A. A. Pedrasa, “Utility-scale DG planning using location-
specific hosting capacity analysis,” in 2016 IEEE Innovative Smart Grid Technologies -
Asia (ISGT-Asia), 2016, pp. 1–6, doi: 10.1109/ISGT-Asia.2016.7796519.
[95] M. S. S. Abad, J. Ma, D. Zhang, A. S. Ahmadyar, and H. Marzooghi, “Probabilistic
Assessment of Hosting Capacity in Radial Distribution Systems,” IEEE Trans. Sustain.
Energy, vol. 9, no. 4, pp. 1935–1947, 2018, doi: 10.1109/TSTE.2018.2819201.
[96] B. Azibek, A. Abukhan, H. S. V. S. K. Nunna, B. Mukatov, S. Kamalasadan, and S.
Doolla, “Hosting Capacity Enhancement in Low Voltage Distribution Networks:
Challenges and Solutions,” in 2020 IEEE International Conference on Power
Electronics, Smart Grid and Renewable Energy (PESGRE2020), 2020, pp. 1–6, doi:
10.1109/PESGRE45664.2020.9070466.
[97] A. K. Jain, K. Horowitz, F. Ding, N. Gensollen, B. Mather, and B. Palmintier, “Quasi-
Static Time-Series PV Hosting Capacity Methodology and Metrics,” in 2019 IEEE
Power and Energy Society Innovative Smart Grid Technologies Conference, ISGT
2019, 2019, pp. 1–5, doi: 10.1109/ISGT.2019.8791569.
[98] H. Al-Saadi, R. Zivanovic, and S. F. Al-Sarawi, “Probabilistic Hosting Capacity for
Active Distribution Networks,” IEEE Trans. Ind. Informatics, vol. 13, no. 5, pp. 2519–
107
2532, 2017, doi: 10.1109/TII.2017.2698505.
[99] S. M. Ismael, S. H. E. Abdel Aleem, A. Y. Abdelaziz, and A. F. Zobaa, “Distributed
generation in deregulated energy markets and probabilistic hosting capacity decision-
making challenges,” in Decision Making Applications in Modern Power Systems,
2020th ed., S. H. E. A. Aleem, A. Almoataz Youssef, A. F. Zobaa, and R. Bansal, Eds.
125 London Wall, London EC2Y 5AS, United Kingdom: Elsevier, 2020, pp. 223–246.
[100] M. J. Chihota and B. Bekker, “Modelling and Simulation of Uncertainty in the
Placement of Distributed Energy Resources for Planning Applications,” in 2020
International Conference on Probabilistic Methods Applied to Power Systems, PMAPS
2020 - Proceedings, 2020, pp. 1–6, doi: 10.1109/PMAPS47429.2020.9183657.
[101] M. J. Chihota and B. Bekker, “New planning principles for distribution networks with
penetration of distributed energy resources,” in 6th IEEE International Energy
Conference, ENERGYCon 2020, 2020, pp. 1–6, doi:
10.1109/ENERGYCon48941.2020.9236463.
[102] D. Fang et al., “Deterministic and Probabilistic Assessment of Distribution Network
Hosting Capacity for Wind-Based Renewable Generation,” in 2020 International
Conference on Probabilistic Methods Applied to Power Systems, PMAPS 2020 -
Proceedings, 2020, pp. 1–6, doi: 10.1109/PMAPS47429.2020.9183525.
[103] F. Ebe, B. Idlbi, J. Morris, G. Heilscher, and F. Meier, “Evaluation of PV hosting
capacity of distribuion grids considering a solar roof potential analysis - Comparison of
different algorithms,” in 2017 IEEE Manchester PowerTech, Powertech 2017, 2017,
pp. 1–6, doi: 10.1109/PTC.2017.7981017.
[104] B. B. Navarro and M. M. Navarro, “A comprehensive solar PV hosting capacity in MV
and LV radial distribution networks,” in 2017 IEEE PES Innovative Smart Grid
Technologies Conference Europe (ISGT-Europe), 2017, vol. 2018-Janua, pp. 1–6, doi:
10.1109/ISGTEurope.2017.8260210.
[105] A. Ballanti and L. F. Ochoa, “On the integrated PV hosting capacity of MV and LV
distribution networks,” in 2015 IEEE PES Innovative Smart Grid Technologies Latin
America (ISGT LATAM), 2016, pp. 1–5, doi: 10.1109/ISGT-LA.2015.7381183.
[106] SA Power Networks, “Technical standard - TS129: small EG connections-capacity not
exceeding 30kW,” Keswick, South Australia, 2019. [Online]. Available:
http://www.sapowernetworks.com.au/centric/industry/contractors_and_designers/techn
108
ical_standards.jsp.
[107] M. Altin, E. U. Oguz, E. Bizkevelci, and B. Simsek, “Distributed generation hosting
capacity calculation of MV distribution feeders in Turkey,” in IEEE PES Innovative
Smart Grid Technologies Conference Europe, 2015, vol. 2015-Janua, no. January, pp.
1–7, doi: 10.1109/ISGTEurope.2014.7028776.
[108] C. Eid, J. Reneses Guillén, P. Frías Marín, and R. Hakvoort, “The economic effect of
electricity net-metering with solar PV: Consequences for network cost recovery, cross
subsidies and policy objectives,” Energy Policy, 2014, doi:
10.1016/j.enpol.2014.09.011.
[109] P. Vilaça Gomes et al., “Technical-economic analysis for the integration of PV systems
in Brazil considering policy and regulatory issues,” Energy Policy, 2018, doi:
10.1016/j.enpol.2018.01.014.
[110] D. C. Matisoff and E. P. Johnson, “The comparative effectiveness of residential solar
incentives,” Energy Policy, 2017, doi: 10.1016/j.enpol.2017.05.032.
[111] I. Koumparou, G. C. Christoforidis, V. Efthymiou, G. K. Papagiannis, and G. E.
Georghiou, “Configuring residential PV net-metering policies – A focus on the
Mediterranean region,” Renew. Energy, vol. 113, 2017, doi:
10.1016/j.renene.2017.06.051.
[112] R. Pacudan, “The economics of net metering policy in the philippines,” Int. Energy J.,
vol. 18, no. 3, 2018.
[113] N. R. Darghouth, G. Barbose, and R. Wiser, “The impact of rate design and net
metering on the bill savings from distributed PV for residential customers in California,”
Energy Policy, vol. 39, no. 9, pp. 5243–5253, Sep. 2011, doi:
10.1016/j.enpol.2011.05.040.
[114] J. Thakur and B. Chakraborty, “Sustainable net metering model for diversified India,”
2016, doi: 10.1016/j.egypro.2016.06.139.
[115] Interstate Renewable Energy Council (IREC), “IREC model net-metering rules 2009,”
Interstate Renewable Energy Council (IREC), United states of America, 2009. [Online].
Available: http://www.irecusa.org/NMmodel09.
[116] O. Zinaman, A. Aznar, C. Linvill, N. Darghouth, T. Dubbeling, and E. Bianco, “Grid-
connected distributed generation: compensation mechanism basics,” Golden, CO
109
80401, 2017. [Online]. Available: https://sam.nrel.gov.
[117] O. Zinaman, T. Bowen, and A. Aznar, “An overview of behind-the-meter solar-plus-
storage regulatory design: Approaches and case studies to inform international
applications (Report no. NREL/TP-7A40-75283),” California, 2020. [Online]. Available:
www.nrel.gov/publications.
[118] A. Satchwell, A. Mills, and G. Barbose, “Regulatory and ratemaking approaches to
mitigate financial impacts of net-metered PV on utilities and ratepayers,” Energy Policy,
2015, doi: 10.1016/j.enpol.2015.05.019.
[119] M. Pesaran H.A, P. D. Huy, and V. K. Ramachandaramurthy, “A review of the optimal
allocation of distributed generation: Objectives, constraints, methods, and algorithms,”
Renewable and Sustainable Energy Reviews, vol. 75. pp. 293–312, 2017, doi:
10.1016/j.rser.2016.10.071.
[120] G. Crabtree et al., “Integrating renewable electricity on the grid,” in AIP Conference
Proceedings, 2011, vol. 1401, pp. 387–405, doi: 10.1063/1.3653865.
[121] A. M. Vassallo, “Applications of batteries for grid-scale energy storage,” in Advances in
Batteries for Medium and Large-Scale Energy Storage: Types and Applications, 2015.
[122] P. F. Ribeiro, C. Z. S. Antonio, and R. Yuri R, Energy Storage at Different Voltage
Levels: Technology, integration, and market aspects. Institution of Engineering and
Technology (IET), 2018.
[123] M. C. Lott and S.-I. Kim, “IEA Technology Roadmap: Energy storage,” Energy Technol.
Perspect., p. 64, 2014, doi: 10.1007/SpringerReference_7300.
[124] B. K. Gardiner, “Applications and Effects on a Medium Voltage Grid,” Delft University of
Technology, 2016.
[125] S. B. Pienaar, K. Kusakana, and P. T. Manditereza, “Usage of Battery Energy Storage
Systems to Defer Substation Upgrades,” 2018, doi: 10.1109/OI.2018.8535635.
[126] SolarEdge, “SolarEdge Export Limitation Application Note,” SolarEdge, 2017. [Online].
Available: www.solaredge.com.
[127] ABB, “ABB solar inverters export limitation,” ABB, 2019. [Online]. Available:
https://library.e.abb.com/public/eb0b2c5947de424a89d1e7871338e7ad/ABB Export
limitation solution.pdf.
110
[128] WesternPower, “Integration Guideline Inverter Embedded Generation,” Australia, 2016.
[Online]. Available: https://westernpower.com.au/media/3403/network-integration-
guideline-inverter-embedded-generation-20190802.pdf.
[129] F. Gonzalez-Longatt and J. Rueda, PowerFactory Applications for Power System
Analysis. 2014.
[130] Interstate Renewable Energy Council (IREC), “IREC model net-metering rules 2005,”
Interstate Renewable Energy Council (IREC), United States of America, 2005. [Online].
Available: http://www.irecusa.org/index.php?id=88.
[131] Erongo RED, “Erongo RED tariff-booklet for 2015 and 2016,” Walvis Bay, 2015.
[Online]. Available: https://www.erongored.com/wp-
content/uploads/2016/06/ErongoRed_annualreport_2016_FINAL_low-res.pdf.
[132] The World Bank, “Namibia - official exchange rate (LCU Per US$, period average),”
2019. https://tradingeconomics.com/namibia/official-exchange-rate-lcu-per-us-dollar-
period-average-wb-data.html (accessed May 26, 2020).
[133] Erongo RED, “Erongo RED tariff-booklet for 2018 and 2019,” Walvis Bay, 2018.
[Online]. Available: https://www.erongored.com/wp-
content/uploads/2018/09/8041_Erongo-RED-Tariff-booklet-18-FINAL-sml.pdf.
[134] K. Palani, R. Kota, A. P. Azad, and V. Arya, “Blue skies: A methodology for data-driven
clear sky modelling,” in Proceedings of the Twenty-Sixth International Joint Conference
on Artificial Intelligence (IJCAI-17), 2017, pp. 3777–3783, doi: 10.24963/ijcai.2017/528.
[135] D. Yang, P. Jirutitijaroen, and W. M. Walsh, “The estimation of clear sky global
horizontal irradiance at the equator,” Energy Procedia 2, vol. 25, pp. 141–148, 2012,
doi: 10.1016/j.egypro.2012.07.019.
[136] M. J. Reno, C. W. Hansen, and J. S. Stein, “Global Horizontal Irradiance Clear Sky
Models: Implementation and Analysis,” Livermore, California, 2012. doi:
10.2172/1039404.
[137] Folsomlabs, “HelioScope: Advanced Solar Design Software.” Folsomlabs, San
Francisco, 2011, [Online]. Available: www.helioscope.com.
[138] Electricity Control Board (ECB), Standard Specification for Medium and Low Voltage
Electricity Distribution Works: Part B-07 - service connections. Windhoek, pp. 1–16.
111
APPENDICES
APPENDIX A
DPL SCRIPT’S SUMMARY RESULTS FOR !NARA RESIDENTIAL DISTRIBUTION
NETWORK (TEST 1)
Script’s Summary Results for July 2018
Script’s Summary Results for August 2018
Script’s Summary Results for September 2018
Script’s Summary Results for October 2018
112
Script’s Summary Results for November 2018
Script’s Summary Results for December 2018
Script’s Summary Results for January 2019
Script’s Summary Results for February 2019
113
Script’s Summary Results for March 2019
Script’s Summary Results for April 2019
Script’s Summary Results for May 2019
Script’s Summary Results for June 2019
114
APPENDIX B
DPL SCRIPT’S SUMMARY RESULTS FOR !NARA NETWORK (TEST 2)
Script’s Summary Results for July 2018
Script’s Summary Results for August 2018
Script’s Summary Results for September 2018
Script’s Summary Results for October 2018
115
Script’s Summary Results for November 2018
Script’s Summary Results for December 2018
Script’s Summary Results for January 2019
Script’s Summary Results for February 2019
116
Script’s Summary Results for March 2019
Script’s Summary Results for April 2019
Script’s Summary Results for May 2019
Script’s Summary Results for June 2019
117
APPENDIX C
TYPICAL LOADING AND VOLTAGE PROFILE OF !NARA NETWORK DURING A VIOLATED PERMORMANCE METRIC (TEST 1)
(JUNE 2019 at 14% MAX GSF)
Voltage Profile Loading Profile
118
APPENDIX D
TYPICAL LOADING AND VOLTAGE PROFILE OF !NARA NETWORK DURING A VIOLATED PERMORMANCE METRIC (TEST 2)
(JUNE 2019 at 29% MAX GSF)
Voltage Profile Loading Profile
119
APPENDIX E
TYPICAL VOLTAGE PROFILES FOR SPAR NETWORK (TEST 3)
(JUNE 2019 at 100% MAX GSF)
….
120
APPENDIX F
TYPICAL LOADING PROFILE OF TRANSFORMER AND CONSUMER FEEDER CABLES (TEST 3)
(JUNE 2019 at 100% MAX GSF)
