Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Algodec (latus sensus)in energy planning -
a (bit biased) review and some challenges
Carlos Henggeler AntunesUniversity of Coimbra and R&D Unit INESC Coimbra
Algorithmic Decision Theory WorkshopUniversity of Manchester, April 2011
Since the early days of Operational Research, the application ofthe models and methods of OR has revealed a very effectivecontribution to the successful resolution of several problems in theenergy sector.
A cross-fertilization has occurred in the sense that the challengingdiversity and complexity of the problems arising in the energysector have fostered new methodological developments to tacklethem in innovative ways that sometimes could be replicated in oradapted to other fields of application.
Introduction
The energy sector is undergoing important changes.
The shift towards the liberalization of the energy markets, namelyin generation, wholesale trading and retailing.
The exigency for sustainable development, balancing economic,environmental and social goals.
Decisions of distinct nature (policy, planning and management) tobe made by different entities (utilities, regulator, governments)must take into account several conflicting objectives (technical,socio-economic, environmental) at various levels of decisionmaking (ranging from the operational to the strategic).
Introduction
Broad application areas
Energy policy analysis- guiding the development and formulation of energy
policies: national/regional energy systems assessment, debateon energy policy, conservation strategies, resource allocation, ...
Electric power planning- strategic planning: power generation expansion
planning, electrical transmission network expansion planning,power distribution planning, ...
Technology choice and project evaluation- evaluation and selection of energy technologies
appraisal of investment projects, ...
Broad application areas
Energy utility operations and management- operational issues in energy industry: biding and pricing,
power plant sitting, energy companies management, ...
Energy-related environmental policy analysis- at policy level: assessment of climate policy, debate on
GHG mitigation are air pollution control policies
Energy-related environmental control and management- waste storage and management, EIA related to major
development projects
Important issues at stake
Complexity- inter-related problems, combinatorial nature, multiple
stakeholders with conflicting views, ...
Uncertainty- extended time frames; data is scarce, controversial,
difficult to obtain; “structural uncertainty”, ...
Multiple criteria- Cost, environmental impacts, reliability, public
acceptance, quality of service, ...
Examples of optimization approaches
Long-term/strategicPower generation expansion planningTransmission network expansion planning
OperationalGeneration schedulingReactive power planningDSM planning
Sort-termUnit commitmentPower flow
Examples of optimization approaches
Power generation capacity expansion planning
Determine the number and type (primary energy source,conversion technology) of generating units and power output tobe installed throughout a planning period.
Minimize costs, minimize pollutant emissions, maximizereliability/safety of supply, minimize external dependence,minimize risk/damage potential, minimize radioactive wastes,...
Constraints: demand (+ reserve margin) satisfaction,capacity bounds, domestic fuel quotas, operational availability,rate of growth of additional capacity, committed power, ...
Examples of optimization approaches
Transmission and distribution network planning
Determine the location, time frame of new lines andother equipment to be installed throughout a planning period.
Minimize costs, minimize population exposure toelectromagnetic fields, minimize visual impact, minimizepotential damage to ecosystems, maximize reliability, minimizebuss voltage deviations, ...
Constraints: meet demand, satisfy operational/technicalrequirements (thermal, voltage drop), power injections, ...
Examples of optimization approaches
Reactive power compensation planning
Determine the number, location and size of devices(shunt capacitors) to be installed in the network.
Minimize costs, minimize power losses, minimize voltagedeviation w.r.t. nominal values (QoS),...
Constraints: powerflow, voltage profile, ...
Examples of optimization approaches
Unit commitment and dispatch
Determine the generation schedule (allocation ofgeneration to the units).
Minimize costs, minimize pollutant emissions, minimizesystem transmission losses, ...
Constraints: bus voltage profile, line overloading,capacity, operational, reserve schedule, ...
Examples of optimization approaches
Load management
Determine load shedding patterns to be applied togroups of end-use loads
Minimize peak power, minimize discomfort caused tocustomers, maximize profits, maximize consumer billreduction,...
Constraints: operational, quality of service, ...
Examples of optimization approaches
Remote load controlDemand at PT2: without(thin) and with (dotted) power curtailments
0
100
200
300
400
500
600
700
800
0:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:0010
:0011
:0012
:0013
:0014
:0015
:0016
:0017
:0018
:0019
:0020
:0021
:0022
:0023
:00
kW
0:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
G 5
G 4
G 3
G 2
G 1 4
G 1 5
G 1 6
G 1 3
G 1 2
G 1 0
G 1 1
G 9
G 8
G 7
G 1 7
G 1 8
G 1 9
G 2 0
G 6
G 1
Examples of optimization approaches
Algorithmic tools- (Multi-Objective) (Mixed Integer) Linear Programming- Goal Programming- (MO) Non-Linear Programming- (MO) Genetic/Evolutionary Algorithms- (MO) Stochastic Programming- (MO) Fuzzy Programming- (MO) Meta-Heuristics
* Tabu Search* Simulated Annealing* PSO* GRASP* ...
Examples of discrete choice problems
Comparative evaluation of power generation technologies- prioritizing technological options- clarifying opposing views in public debate (nuclear
option vs. conventional thermal generation)- comparing between different renewable options to
derive priorities complying with broad policy goals.
Selection between alternative energy plans and policies- choice between alternative strategies at
regional/national level using scenarios (renewables, biomasscrops, ...)
- group decision and negotiation in face of severalstakeholders
Examples of discrete choice problems
Sorting out candidate energy projects- DSM options- alternative plans for network expansion or
reinforcement- classification of DSM and supply options in groups
according to attractiveness- portfolio approaches
Sitting and dispatching decisions- thermal or nuclear power plants- corridors for transmission lines- dispatching in face of disturbances threatening the
system’s stability and security
Stakeholders involvement / Structuring
“A well structured problem is a problem half solved”.
Address all types of concerns from multiple stakeholders thatshould be encompassed in evaluation models.
Better understand the conditions in which the final solution willbe implemented.
Use of SSM to structure the initial ‘messy situation’ and to helpunveiling a ‘‘cloud of objectives” for a multi-criteria evaluation ofenergy efficiency initiatives.
Refinement of the ‘‘cloud of objectives” and development oftrees of fundamental objectives based on Keeney’s ValueFocused Thinking guidelines.
Demand-Side Management (DSM) has been recognized as aneffective tool for increasing the energy efficiency of the economyand reducing the environmental impact of energy use.
Utilities have been stimulated through regulation to promote DSMwith financial compensations.
Market Transformation: change the market on a permanent basis,reducing the barriers to the natural adoption of energy efficiencyas a criterion of equipment choice or everyday practice by end-users.
MCDA for sorting energy efficiency initiatives, promoted by electricutilities (generally with public funds authorized by a regulator).
Evaluation of actions for promoting end-use energy efficiency
Structuring phase using SSM: identifying the main actors, theirpoints-of-view regarding energy efficiency and extending theknowledge about the problem.
Entities that could be interested in using this evaluation system:• Energy Agency• Regulator• Distribution utilities• competitive Supply companies (although naturally aiming
at increasing sales, may face energy efficiency as a marketingstrategy to attract or to keep customers).
Value-Focused Thinking
Evaluation of actions for promoting end-use energy efficiency
MCDA advantage- inclusion of impacts usually not considered due to the
difficulty or impossibility of being measured in monetary units,- enabling the DM to base his/her decision on his/her own
values, instead of using the conversion rules hidden in themonetization formulae.
Provide more confidence in the decision suggested, also due to- the absence of compensation effects (a good performance
in one criterion does not hide a poor performance in another),- the possibility of assessing the robustness of the
decisions regarding the uncertainty of the input data.
Evaluation of actions for promoting end-use energy efficiency
Evaluation of actions for promoting end-use energy efficiency
Tree of fundamental objectivesof an energy agency
Tree of fundamental objectivesof a regulator
Evaluation of actions for promoting end-use energy efficiency
Ref. Title Description
a1 Load management for
commercial clients.
Installation of a load controller for peak cutting and load shifting in commercial consumers,
complemented with education through seminars.
a2 Improvements in
manufacturing processes.
Industrial engineering support and financial incentives to allow customers and utility staff to
explore specialized industrial energy savings opportunities, complementing rebate
programmes.
a3 Industrial Power Smart:
Employee involvement.
Incentive to industrial employees, for identifying energy-efficiency measures with the aim of
acquiring low-cost savings. The programme is promoted on the industrial customers and
seminars are offered to the employees, which receive a monetary incentive for each efficiency
action suggested and for the effective savings.
a4 Industrial Power Smart:
Compressed air
component.
Detailed study of the participant's compressed air system, action plan and financial
assistance.
a5 Efficient lighting for
schools.
Performance contracting for a school building, aiming at energy saving measures for an
efficient illumination system for schools (Pilot Project).
a6 Bonus for savings above
15%.
Consumers that save more than 15% of their annual electricity use get a bonus of 50 Euro.
Information about energy savings is provided to participants on request.
a7 Promotion of home
appliances with low stand-
by losses.
Subsidies to high efficient home appliances with low stand-by losses or automatic switch off in
the stand-by mode.
a8 Energy management in the
public sector.
Education of directors, technical staff and remaining personnel in the public services through
seminars, and the arrangement of cooperative networks between energy managers of the
public institutions.
a9 Energy management in
buildings with area >
1500m2.
Annual energy audits to big buildings with classification regarding energy consumption and a
mandatory efficiency measures planning.
a10 Washing at lower
temperatures.
A marketing campaign with the purpose of reducing the number of laundry washes above
60ºC.
a11 Energy consultancy for
industries with energy
consumption above 2
GWh/year.
Free audits conducted in big industrial consumers which can apply for external subsidies
regarding measure installation costs.
a12 Night rate campaign. Campaign for night rate tariff supporting electricity use in off-peak hours.
a13 Heat storage with night
time rates.
Introducing accumulated hot water and heating storage systems in the residential sector
through rebates.
a14 Variable Speed Drives
(VSD) and efficient motors.
Promotion of electronic speed regulation of engines or the replacement of old motors by high
efficiency units.
a15 Heat pumps. Promotion of heat pumps for domestic space heating.
a16 Efficient lighting in SMEs. Promotion of high efficiency lighting systems for Small and Medium size Enterprises (SMEs).
a17 Domotics. Installation of consumption search equipments to rationalize the electric consumption in the
domestic sector, improving general comfort.
a18 Promotion of A and B label
fridges.
Rebates in domestic fridges of efficiency classes A and B to make them more attractive to
consumers (minimization of the initial cost difference to lower efficiency models).
a19 High efficiency motors. Promoting high efficiency motors for industries
a20 Public lighting efficiency
improvements.
Installation of regulation and/or replacement with more efficient components.
a21 Combined DSM actions. Marketing campaigns and rebates for the domestic and commercial sectors on two specific
geographic areas: 1) of predominating residential loads (55%), and 2) of predominant
commercial loads with the purpose of saving energy and peak demand.
a22 Compact Fluorescent Light
bulbs (CFLs) paid back
through the bill.
Dissemination of CFLs in the residential sector by supplying bulbs to residential consumers
which will be paid back through the differences in the electricity bill.
a23 Low flow shower heads. Promotion through rebates of low flow shower heads to consumers with electric water heating
systems.
a24 Cool storage. Promotion of cool storage systems for commercial buildings.
Ref. Participants Useful life Energy savings Peak savings Part. cost Promoter cost Total cost
(years) MWh MW (103 Euro) (10
3 Euro) (10
3 Euro)
a1 6 10 2592 67.5 5330 17780 23110
a2 517 10 390025 29.3 12408 4653 17061
a3 15 8 4080 0.1 0 251 251
a4 181 10 65703 9.9 3391 3567 6958
a5 1 10 270 0.0 2 66 68
a6 150 10 540 0.0 16 8 24
a7 250 10 80 0.0 0 8 8
a8 700 5 197750 4.5 6653 2069 8722
a9 2500 10 200000 2.3 5887 4701 10588
a10 279586 10 139793 16.0 0 977 977
a11 12 5 79326 1.8 0 1864 1864
a12 54736 10 0 61.0 17682 5474 23156
a13 1872 10 0 3.7 0 1471 1471
a14 7 10 15130 0.3 0 55 55
a15 156 10 76800 7.2 521 368 889
a16 77330 10 98980 1.2 782 644 1426
a17 252 10 7050 0.9 151 50 201
a18 6898 10 18936 0.2 472 194 666
a19 83688 10 1081500 18.2 2667 750 3417
a20 30000 10 107102 2.5 479 251 730
a21 3870 8 12508 1.2 529 461 990
a22 60000 6 16200 0.0 316 61 377
a23 50000 5 15000 1.0 77 27 104
a24 100 10 0 25.0 162 6700 6862
Description of the alternatives
Performance of eachalternative in each criterion
Evaluation of actions for promoting end-use energy efficiencyELECTRE TRI methoddevoted to the sortingproblem: assigning eachalternative to one of a set ofpre-defined orderedcategories according to aset of evaluation criteria.Categories are defined byspecifying their boundariesby means of referenceprofiles, in terms ofperformance in eachcriterion.
Energy planning: complex technological systems interacting inmultiple ways with economic, social and natural environment.
Targeting for a more or less distant future, for which forecast aredifficult to be made: oil prices, inflows into a reservoir, lack ofhuman experience with some phenomena, etc.
Internal: problem structuring and elicitation of values.
External: limited knowledge about the magnitude and evolution ofimportant parameters.
Construction of scenarios, sensitivity analysis, qualitative scales,stochastic / fuzzy / interval / rough set approaches, …
Uncertainty
Data Envelopment Analysis (DEA)- Methodology devoted to frontier analysis
- Uses empirically available information - Non-parametric technique: not requiring the a-prioriimposition of any specific functional form (e.g., regressionequation, production function) relating independent variables(inputs) with dependent variables (outputs) - Efficiency in the Pareto-Koopmans sense - DMUs are expected to operate in a relativelyhomogeneous environment
- DMUs should possess some management autonomy.
Performance evaluation
Projection mechanism: DEA models determine the projections of the inefficientDMUs on the efficiency frontier.
Projection of an inefficient DMU obtained through a linearcombination of the efficient DMUs that define the face of theenveloping surface containing the projected point.
Performance evaluation
(Xk,Yk) DEA Model (X,Y) (Xk,Yk)
Performance evaluation
- Electricity distribution companies
- Power plants
- Generation technologies
- International comparisons
- Gas distribution companies
Performance evaluation
Most common inputs - Operational costs (Opex) - Capital costs (Capex) - Maintenance costs - Labour (# employees, labour hours) - Labour wages (administrative, technical) - Maximum peak load (proxy for transformer capacity) - Purchased power
- T&D Losses (Joule effect in lines) (proxy for technical quality) - Transformer capacity (MV, HV) - Network length (Km) (also a proxy for capital stock)
buried and aerial per voltage level (high, medium, low)
Performance evaluation
Most common outputs - Network length (Km) buried and aerial per voltage level
(LV, MV, HV) - # total customers (residential, industrial)
- # customers per activity sector (industry, services, residential)- Energy distributed to end-users (units sold KWh) (by sector)
- Energy sold to other utilities - Peak power (MW) - Profits - Energy quality (SAIDI, SAIFI, frequency, waveform, ...) - 1/losses (proxy for technical quality of the grid) - Inverse density index (settled area in Km2 per inhabitant) [improves the performance of sparsely inhabited distribution areas]
Performance evaluation
Most common environmental factors (uncontrollable variables,non-discretionary, exogeneously fixed) - Network length (Km) - # clients - Customer density (#/Km2) - Geographical dispersion - Load factor - Weather (Winter conditions, snow) - Forest area - Other particularities (West vs East Germany, rural vs urban)
Outside the control of the management (can be regarded asgiven)
Performance evaluation
Use of DEA for regulatory purposesBenchmarking has become a widely used tool in
incentive regulation of utilities. The aim of incentive regulation is to promote efficiencyimprovements in the absence of market mechanisms.
Generation and retail are potentially competitive. Transmission and distribution are subject to regulation.
Incentive schemes to promote cost saving, investmentefficiency and service quality.
Incentive regulation schemes for QoS have laggedbehind schemes for achieving cost efficiency.
Under the prevalent regulation schemes, utilities facestrong incentives to undertake cost savings.
Performance evaluation
QoS comes at a cost!Do companies respond to cost saving incentives by reducingservice quality rather than pursuing real efficiencyimprovements?Challenge: to maintain well-balanced financial and qualityindicators. Information and Incentives Project (IIP), UK, 2002/03 - defined output measures for service quality - linked the quality performance of the DNOs to theallowed revenue: * penalize utilities for not meeting the target * reward utilities that exceed the target * reward frontier performance by guaranteeing lessstrict standards for the next control period
Performance evaluation
The Finnish regulator Energy Market Authority (EMA) used DEAas the benchmarking method in regulation.
Input- OPEX (actual operational costs)
Output- Power quality- Value of delivered energy
Environmental factor- Length of network- Number of customers
Reasonable operational costs: RC = (DEA score + 0.1)*OPEX
Performance evaluation
DEA and MCDA as complementary tools
Case Study: Efficiency Benchmarking of Agricultural BiogasPlants in Austria
Representative set of 41 energy crop digestion plants in Austria
Factors used in this study: (1) Labor input (time) (2) Organic dry substance (ODS) input
(3) Biogas or net electricity produced(4) GHG (undesirable output)
Performance evaluation
4 efficiency categories were defined to classify the DMUsaccording to their efficiency:C1 = “Poor”, C2 = “Fair”, C3 = “Good”, and C4 = “Very good”.
To maximize:g1 – Electricity / Labourg2 – Electricity / ODSg3 – Heat / Labourg4 – Heat / ODS
To minimize:g5 – GHG / Labourg6 – GHG / ODS
Performance evaluation
Category definitions for each indicator:
Categoryg1 (max)Elec./Labor
g2 (max)Elec./ODS
g3 (max)Heat/Labor
g4 (max)Heat/ODS
g5 (min)GHG./Labor
g6 (min)GHG./ODS
C1 - Poor < 580 < 960 < 150 < 130 > 250 > 210
C2 - Fair [580, 1 100[ [960, 1130[ [150, 375[ [130, 530[ ]130, 250] ]155, 210]
C3 - Good [1100, 2300[ [1130, 1300[ [375, 950[ [530, 880[ ]80, 130] ]90, 155]
C4 - Very good
≥ 2 300 ≥ 1 300 ≥ 950 ≥ 880 ≤ 80 ≤ 90
If the DM seeks more discriminative results, the performanceranges can be partitioned into a larger number of categories.
Performance evaluation
Possible to add meaningful preference information:e.g., the most important output is electricity, followed by
GHG emissions (to be minimized), and lastly by heat:- wg1 (Electricity/Labor) > wg5 (GHG/Labor) > wg3 (Heat/Labor)
- wg2 (Electricity/ODS) > wg6 (GHG/ODS) > wg4 (Heat/ODS).
and ODS is more important than labor:- wg2 (Electricity/ODS) > wg1 (Electricity/Labor)- wg4 (Heat/ODS) > wg3 (Heat/Labor)- wg6 (GHG/ODS) > wg5 (GHG/Labor)
Sustainable production of electricity and transportation fuels.
Biomass conversion, biofuels and bioenergy.
Photovoltaic solar energy.
Wind and wave energy.
Fuel cells.
Interdisciplinary approaches for sustainable energy technologyassessment include a fundamental engineering analysiscomponent.
Energy and sustainability
Adding intelligence to all areas of the electric power system tooptimize the use of electricity.
Benefits include improved response to power demand, moreintelligent management of outages, better integration of renewablesources, and the storage of electricity.
Ability to sense, monitor, and control (automatically or remotely)how the system operates or behaves under a given set ofconditions.
Using ICT to improve the electricity “supply chain” from powerplants to consumers, allowing consumers to interact with the grid,and integrating new technologies into the operation of the grid.
The smart grid challenge
The energy sector is of outstanding importance for thesatisfaction of societal needs, providing directly orindirectly the fundamental requirements for mostactivities involving human beings from comfort totransportation and production systems.
Ill-structured contexts characterized by technologicalevolution, changes in market structures and new societalconcerns.
Multiple decision agents (government, regulators,utilities, consumers) and evaluation criteria (economic,technical, environmental): grasping the trade-offs.
Conclusions