SINTEF Energy Research

N. Chiesa, M. Korpås, O. E. Kongstein, A. Ødegård

1

Outline1. Introduction

2. Electrical system & Simulation model

3. Dynamic regulation of hydrogen production

4. Case studies

5. Conclusion

Dynamic control of an electrolyser for voltage quality enhancement

The research leading to these results has received funding from the Fuel Cells and Hydrogen Joint Undertaking under grant agreement n° 245262 – NEXPEL

SINTEF Energy Research

NEXPEL main objective: Develop and demonstrate a PEM water electrolyser integrated with RES:

75% Efficiency (LHV), H2 production cost ~ €5,000 / Nm3h-1, target lifetime of 40,000 h

New membrane materials

New catalysts Improved MEAs

Novel stack design and newconstruction materials

Improved DC-DC converter

Jan 2010 - Dec 2012Co-ordination: SINTEF

Funding: Fuel Cells and Hydrogen JUTotal Budget: € 3,353,549

www.nexpel.eu

SINTEF Energy Research

• Hydrogen production attractive for integration in wind turbine systems

• Re-conversion is economically challenging

• Hydrogen used locally

• “Smart” operation of electrolyser

• Improvement of power quality at PCC

• Reduction of system losses

3

Hydrogen Production from RES

Demonstrate the feasibility and advantages achievable from the integration of an

electrolyser system for the production of hydrogen in a renewable energy system (RES)

SINTEF Energy Research

• Fluctuating power output of RES influence the operation of electrolyser.

• Large atmospheric alkaline electrolyser: due to long response time (several minutes) are designed to operate at constant power.

• Pressurized alkaline electrolyser: faster response time, current interruption leads to increased degradation rate (min load 25-50%).

• Polymer electrolyte membrane (PEM) electrolyser: fast response time, no degradation with stop-start cycles, higher energy efficiency. Promising but immature technology, expected life time of 10 years.

4

Electrolyser Technology

“classical operation”: power fluctuations to grid, constant

power to electrolyser

“smart operation”: electrolyser absorbs fast fluctuations, grid

receives smooth power electrolyser with flexible

operating capabilities

SINTEF Energy Research

• Simplified representation of a possible island system: wind turbine and electrolyser connected to a relatively weak grid

5

Electrical System

ELYH2

DCAC

40 km 30 km Zgrid

0.690/22 kV2.5 MVA

SCIG2 MVA

0.690/22 kV625 kVA625 kW500 kW

Cable Cable22 kV50 Hz

Sk = 5 MVAR/X = 1.5

Sk = 9 MVAR/X = 0.67

Kaimal Wind Model

Aerodynamic Turbine Model

B A

Best wind resources often found in areas with weak grid connection to the main

transmission grid: voltage variation and thermal limits may put a significant limit on the realizable wind power generation.

Representative for several location along the Norwegain coast: high wind speed, low

local electricity demand. Hydrogen from electrolysis can be considered for local

and sea transport.

SINTEF Energy Research

• Stochastic wind speed: Kaimal model

• Average wind speed = 7.5 m/s

• Standard deviation = 1.0 m/s

• 60 s window

• General wind turbine aerodynamic torque model with 3P effect

• Squirrel cage induction generator (SCIG)

6

Simulation Model: Wind

ELYH2

DCAC

0.690/22 kV2.5 MVA

SCIG2 MVA

0.690/22 kV625 kVA625 kW500 kW

/X

Kaimal Wind Model

Aerodynamic Turbine Model

B

Main : Graphs

x 10 20 30 40 50 60

(pu)

Shaft speed pu

0.200 0.250 0.300 0.350 0.400 0.450 0.500

(pu)

P wind

SINTEF Energy Research

• Electrolyser: dynamic equivalent model

• Urev: reversible potential of water splitting reaction

• RΩ: ohmic resistance of the cell

• Rct: charge transfer resistance

• Cdl: double layer capacity

• Parameters based on in-house measurements on an alkaline electrolyser

• Electrolyser converter: average model

• Three-phase, two-level PWM converter

• Switching effect phenomena averaged over the switching period

7

Simulation Model: Electrolyser

ELYH2

DCAC

0 690/ 5

SC G

0.690/22 kV625 kVA625 kW500 kW

/

d ode

e ody a c u b e ode

Rct Cdl

RΩ

Urev

SINTEF Energy Research

• Three-phase current reference for the electrolyser converters current conteoller is generated based on active and reactive power references.

8

Dynamic Regulation of the Hydrogen Production

PLL

Qref

Pref N/D N/D

N/D N/D

U_pll Sbase

U_pll Sbase

I_act

I_rea

Generation of current reference

V_meas

U_pll phase

Current Controller

I_meas

I_ref

SINTEF Energy Research

• Electrolyser converter used to compensate the reactive power fluctuation (STATCOM-like)

• Indirect control of the bus voltage: reactive power measured at the generator - Qoff

• Direct voltage control:

• Droop function and load compensation units may be added

9

Dynamic Regulation of the Hydrogen Production

LP filter+ -

Qoff40 ms

Qgen Qref

-Qmax

Qmax

LP filter + -

Vref5 ms

U_pll I_rea

-Imax

Imax + +Gain

1s

Gain

10

1

0.01

SINTEF Energy Research

Pref>Pset

Pref<Pset

LP filter HP filter + + -+ -+

1s

1s

- +

Gain N/D + -

- -

Pset-Pmin

Pely-Pset

Pset

PminPely0

Pely-Pmin

40 ms 1 s

0.5 MW

0.0001

Pgen Pref

Poff Psat

• Strategy used for Q control is not very flexible for P control: constant offset

• Flexible control strategy:

• compensates (fast) active power fluctuation

• allows slow variation

• maximize hydrogen production rate

10

Dynamic Regulation of the Hydrogen Production

Calculation of offset reference signal. Target:

average Pely = Pset

Increase the regulator dynamic response when

maximum regulation ranges are exceeded

SINTEF Energy Research

• Nine case studies with different control strategies:

11

Case Studies

Case E wind E grid E ely

E loss line

E loss Ely

Kg H2 Nm3 H2 E H2

[kWh] [kWh] [kWh] [kWh] [kWh] [kg] [Nm3] [kWh]

1: No Ely 12.3 10.5 0.00 1.78 0.00 0.00 0.00 0.00

2: Ely Max 12.3 3.08 8.23 0.998 3.04 0.16 1.74 5.2

3: Ely Max, Q comp

12.3 3.17 8.15 0.997 3.00 0.16 1.72 5.15

4.1: 100% reserve

12.3 6.33 4.82 1.16 1.58 0.1 1.08 3.24

4.2: 50% reserve

12.3 5.01 6.22 1.08 2.13 0.12 1.37 4.09

4.3: 20% reserve

12.3 3.78 7.53 1.01 2.71 0.14 1.61 4.82

5.1: Const Ely at E_H2 of 4.1

12.3 6.28 4.82 1.21 1.53 0.1 1.10 3.29

5.2: Const Ely at E_H2 of 4.2

12.3 5.02 6.22 1.08 2.12 0.12 1.37 4.09

6: 50% reserve, V control

12.3 4.98 6.2 1.14 2.13 0.12 1.36 4.07

Electrolyser reduces the losses in the line, but high conversion losses:

electrolyser efficiency is crucial

Hydrogen production only slightly affected by best dynamic control

strategies

Dynamic vs. Constant electrolyser power: no increased losses and same

H2 production

Indirect vs. Direct voltage control: almost identical average behavior

SINTEF Energy Research 12

Case Studies

2: Constant maxH2 production

4.1: 100% reserve,PQ ctr, fix offset

4.2: 50% reserve,PQ ctr, dyn. offset

6: 50% reserve,PV ctr, dyn. offset

P

Q

V

+5%

-5%

SINTEF Energy Research

02

46

810

-600

-400

-200

0

200

400

600

800

0 20 40 60 80 100 120 140

600-800

400-600

200-400

0-200

-200-0

-400--200

-600--400

Gai

n di

ffer

ence

[k€/

year

]

• Simplified cost analysis of the wind-electrolyser system with the best control strategies

• The simulated 60 second time period is taken as basis for the economic calculations, by assuming it to be representative for one year

• The total annual wind generation then sums up to 5534 MWh for the 2 MW turbine, corresponding to a capacity factor of 32 %

• Economical estimates based on an electricity price 50 €/MWh, a hydrogen price 5 €/kg and an electrolyser total cost of 5 000 €/Nm3. The margin and the payback time are calculated with reference to the case with no electrolyser. O&M costs are not considered.

• Payback time between 2 to 3 years.

13

Cost Analysis

SINTEF Energy Research

• Demonstration of possible and “smart” use of an electrolyser in a RES

• Voltage quality at PCC is improved by introducing an electrolyser with flexible operating capabilities

• Modelling approach and analysis tools demonstrated in the paper are valuable instruments for the investigation, planning and evaluation of future possibilities for the integration of hydrogen and wind energy technologies

• Economical considerations demonstrate that at today’s electricity prices and expected hydrogen prices, the production of hydrogen from wind energy can become economically feasible

• Can the improved power quality and/or the improved wind energy utilization defend the extra costs of a larger electrolyser required for dynamic control? Other alternatives are flywheel energy storage or reinforcing the local grid

• The effect of dynamic vs. constant load on the electrolyser on the aging rate of the stack need to be further verified (lifetime)

14

Conclusion

SINTEF Energy Research

Technology for a better society

15