New Modellierung und kombinierte Simulation eines Power-to-Gas … · 2015. 7. 10. · Modellierung...

Modellierung und kombinierte Simulation einesPower-to-Gas Prozesses

Andreas Fleischhacker

Technische Universitat WienInstitut fur Energiesysteme und Elektrische Antriebe

Arbeitsgruppe Elektrische Anlagen

13. Februar 2014

1 / 20

Inhaltsverzeichnis

1 Motivation

2 Power-to-Gas ModellModell des alkalischen Elektrolyseurs (AEL)Modell der Methanisierung

3 Kombinierte Simulation

4 Ergebnisse der Analyse

5 Schlussfolgerungen

6 Quellen

2 / 20

Motivation

Das Power-to-Gas Konzept2. Background 2.2. The Power-to-Gas Concept 5

Wind PowerPlant

Photovoltaic

otherrenewables

Electrolysis

CO2 TankMethanation

CHP Plant Pore Storage

H2O O2

H2O

H2

CO2

Electricity Network Natural Gas Network

CO2 Source

Gas to Power

Power to Gas

Power-to-Gas Plant

Figure 2.2.: The concept to store the renewable power in the natural gas grid by trans-forming electrical power into methane. Picture based on [41]

grid mutually. The regenerative generated power can be shifted temporarily with the useof the gas grid.

According to Figure 2.2 the P2G plant connects the electricity network to the naturalgas network. The electrolysis is used to split up water and generate hydrogen, by usingelectricity. This step transformed the electrical into chemical energy. Austria’s hydrogenstructure is sparsely extended, but the 4mol% can be fed into the natural gas grid [35].The next step in the P2G plant is the methanation to combine hydrogen with CO2 andgenerate methane. If the electrical power for this process is generated by renewable powergeneration, the methane is called Renewable Power Methane (RPM) [41].Methane’s advantages are

• storages of methane - the natural gas grid - is available,

• higher energy density of methane to hydrogen and

• the transfer capacities of gas pipelines are much higher than these of electricalgrids.

The synthetic methane can be stored or re-converted to electricity with gas turbinesand/or combined cycle power plants. Another product of the methanation the processheat can be used for district heat, local heat or industrial processes. Besides the technical

3 / 20

Power-to-Gas Modell

Konzept

Nennleistungsskalierung durch die Parameter NS und NP

3. Mathematical Modelling of the Power-to-Gas Process 3.1. Electrolyser 9

consists of the flow rate of hydrogen VH2,AEL(t) and the flow rate of water needed forthe electrolysis VH2O,AEL(t). These output variables are produced by the electrolyser.Another part of the output is generated by the methanation, the flow rate of methanVCH4,meth(t), water VH2O,meth(t), the generated heat Pheat,meth(t) and carbon dioxidefor the methanation VCO2,meth(t). A more detailed explanation of the electrolyser andmethanation pictured in Figure 3.2 will follow in Sections 3.1 and 3.2.To relieve the readability, time dependence of the variables is assumed, but not writtene.g.

xin,P2G(t) = xin,P2G. (3.3)

3.1. ElectrolyserThe electrolyser, the model’s first part transforms electrical energy into the chemicalenergy carrier hydrogen. The beginning of this Section describes the process in theelectrolyser and the types of electrolysers. The second part is a detailed modelof theelectrolysis consisting of an electrochemical and a thermal model.

3.1.1. Types of electrolysersAccording to [36] and [39] the three main types of electrolysers are

• the Alkaline Electrolysers (AEL),

• the Proton Exchange membrane Electrolysers (PEMEC) and

• the high temperature (HTEL) with Solid Oxid Electrolysers (SOEL),

listed up in Table 3.1, containing further characteristics.

Alkaline electrolysers use a porous diaphragm of ceramics (e.g. NiO) or synthetics (e.g.polysulphone) to separate the H2 and O2 flow. The diaphragm is gas-proof, has a lowelectrical resistance and a high mechanical stability. The liquid electrolyte is based ona about 30% wt. KOH solution. Today alkaline electrolysers are commercial available(> 1MW) and long term tested [36].

fAEL

xin,P 2G

yout,AEL =xin,meth

fmeth

Methanationyout,meth

Power-to-Gas plant

Figure 3.2.: Structure of the P2G Model consisting of an electrolyser and a methanation

Electrolyser model

xin,P2G = [Pin, NS , NP ]T

yout,P2G = yout,meth =[VCH4 , VH2O, VCO2

]T

4 / 20

Power-to-Gas Modell Modell des alkalischen Elektrolyseurs (AEL)

Modell des alkalischen Elektrolyseurs (AEL)


the electrolysers output

yout,AEL = VH2 . (3.25)

The two submodels influence each other, e.g. a higher production of hydrogen resultsin a increasing electrolyser temperature. A raising electrolyser temperature decreasesthe voltage of the electrolyser. The electrochemical model describes the relationship ofelectrical input power, produced hydrogen and efficiency. The thermal model calculatethe electrolyser temperature and the cooling demand, depending on the efficiency.

xin,P 2G

fin,AEL

fAEL,ec

fAEL,th

xin,AEL fout,AEL

yout,AEL

yout,AEL,ec

yout,AEL,th

Electrochemical model

Thermal model

Alkaline electrolyser

Figure 3.4.: Structure, in- and outputs of the alkaline electrolyser model

3.1.3. Electrochemical ModelAccording to [43] electrode kinetics of one AEL cell can be modelled by a current-voltagerelationship

Ucell(p,T ) = Urev(p,T ) + r1 + r2 T

Aicell + s log

(t1 + t2

T + t3T 2

Aicell + 1

). (3.26)

The voltage of one cell Ucell is formed by

• the voltages of anode and cathode,

• the ohmic voltage losses,

• the bubble voltage,

• the diaphragm voltage and

• the electrolyte voltage.

xin,P2G = [Pin, NS , NP ]T

yout,AEL = VH2

5 / 20


Elektrochemisches Modell - PV Ansteuerung

Testfunktion: typischer PV Tagesgang + Standbyleistung1 Mathematical Modelling of the Power-to-Gas Process 1.1 Electrolyser 12

00:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

00:00

0.2

0.4

0.6

0.8

1

t in daytime

Pin

MW

Pin

0

0.2

0.4

0.6

0.8

1·107

Vin

Nm

3VH2

Figure 1.7: Produced hydrogen in dependence to the input power Pin. The input equalsthe power of a typical PV production.

20 40 60 80 10060

70

80

90

100

Pin/Pn,AEL in %

η ein

%

Figure 1.8: Efficiency of an alkaline electrolyser over the input power. The electrolysertemperature is constant at ϑ = 85 ◦C.

As a conclusion the output as a function of the electrochemical model fAEL,ec can bewritten by

yout,AEL,ec = fAEL,ec(xin,AEL,yout,AEL,th) =

NS Ucell

PinUAEL

ηFNS NP

z FVstd

UthUcell

. (1.45)

6 / 20


Thermisches Modell

Temperaturabhangigkeit

T = 1CT

(Qgen − Qloss − Qsens − Qcool

)

Regelung durch die Kuhlleistung Qcool

3. Mathematical Model of the Power-to-Gas Process 3.2. Methanation 24

3.2. MethanationMethanation is a standard technology used for coal gasification and developed for biomassgasification [46]. In a P2G plant methanation is used to transform H2 and CO2 to methane(CH4). Reasons [46] to use CH4 (SNG) instead of H2 are

• that the natural gas grid allows a wide feed-in and a huge storage capacity and

• the volume reduction by a factor 5:1.

Methanation was found in 1902 by Sabatier and Senderens. The two main applicationsof the methanation process are

• the removal of CO in hydrogen rich gases and

• the conversion of synthesis gas to synthetic natural gas (SNG) [28].

Three typical reactor types are fixed-bed, fluidised-bed and Slurry Bubble ColumnReactor (SBCR). Table 3.3 shows the advantages and disadvantages of these three types.According to [28] conversion can be described by three reactions, the methanation reaction

3H2 + CO←→ CH4 + H2O, ∆H0R = −206.28 kJ/mol , (3.59)

and the water gas shift

CO + H2O←→ CO2 + H2, ∆H0R = −41.16 kJ/mol. (3.60)

The exothermic reactions (H0R negative) pass through a catalyst build of nickel, platinum,

iron or ruthenium. Nickel is the most common material due to financial and technical

0 1 2 3 4 520

40

60

80

t in hours

ϑin

◦ C

ϑ

0

5

10

15

qin

%

qcool

Figure 3.10.: Temperature step response of the AEL electrolyser in red and the bluestandardised cooling demand qcool required to control the temperature.Starting temperature ϑstart = 25 ◦C.

7 / 20

Power-to-Gas Modell Modell der Methanisierung

Modell der Methanisierung - Triviales Modell

das Gasnetz wird als ideale Senke betrachtetkeine Betrachtung der Dynamik

yout,P2G = yout,meth =[VCH4 , VH2O, VCO2

]T

Reaktion mit Kohlenstoffdioxid

4H2 + CO2 ←→ CH4 + 2H2O∆H0

R = −165.12 kJ/mol

8 / 20

Kombinierte Simulation

Gesamtmodell der P2G AnlageSimuliert in Matlab/Simulink R©

Combined electrochemical and thermal model!Run parameter.m

Stack Outputs

V

Nm^3 per h

Nm^3 per h

T_stack

yellow ... Q_genmagenta ... Q_lossturquoise ... Q_coolred ... Q_sensible

Scopes

V

Nm^3 per h

Nm^3 per h

Production

1

mA per cm^2

kW per Nm^3 H2

kW

A

W

kWh

Cell outputsN_s

standby energy

start at 1s

power

f(u)

consumption

f(u)

Thermal subsystem with Controller

In1

In2

In3

In4

In5

In7

Out1

Out2

Out3

Out4

T test

I_max I_max*f

PV with saturationI_max

noiseI_max*f

PVP_n

noiseP_n * f_PV

dV_H2_dt

dV_CH4_dt

dV_CO2_dt

dV_H2O_dt

dQ_methanation_dt

Methanation

P_in

P_min

U_stack

start

I_in

delta_P

controller

i_in

theta_stack

u_cell

eta_e

dV_H2cell_dt

dV_O2cell_dt

j_cell

cell

1s

1s

1s

1s

1s

1s

[theta_stack]

[j_cell]

[i_in]

[dV_O2_dt]

[dV_H2_dt]

[u_stack]

[eta_e]

[V_H2O]

[dV_O2cell_dt]

[V_CO2]

[V_CH4]

[dQ_cool_dt]

[m_H2O]

[V_O2]

[V_H2]

[delta_P]

[i_P]

[theta_set]

[Q]

[dV_H2cell_dt]

[u_cell]

-K-

1/10

-K-

-K-

-K-

[dV_H2_dt]

[eta_e]

[i_in]

[u_stack]

[j_cell]

[u_stack]

[V_H2O]

[V_CO2]

[V_CH4]

[u_stack]

[i_in]

[u_stack]

[dV_H2_dt]

[theta_set]

[dV_O2cell_dt][dV_H2_dt]

[eta_e]

[V_O2]

[V_H2]

[dV_O2_dt]

[dV_H2_dt]

[i_P]

[i_in]

[delta_P]

[dQ_cool_dt]

[dV_H2cell_dt]

[dV_O2_dt]

[dV_H2_dt]

[theta_stack]

[Q]

[m_H2O]

[theta_stack]

[dV_O2cell_dt]

[dV_H2cell_dt]

[u_cell]

[dV_O2_dt]

[i_in]

SINCAL_P_in

[u_cell]

1

modelparam.P_n*0.2

75

0

modelparam.I_n

-C-

modelparam.P_n*0.95

modelparam.I_n

9 / 20


Impementierung in zwei Modellnetze

I Zwei Sincal R©ModellnetzeI Landliches Netz

Haushalte (H0 und L0) mit PV, Windpark (9MW)I Suburbanes Netz

Haushalte (H0) mit PVI Matlab R©

I AlgorithmusI Optimierung der Speichergroße

I Untersuchung vonI 2 SzenarienI Betriebsmodi der P2G AnlageI dynamische Wirkungsgrade

10 / 20


Szenarien

Zwei Szenarien mit unterschiedlichem PV Ausbaugrad

I Szenario 1:I maximale PV AusbaustufeI maximale Netzbelastung

I Szenario 2:I quasi ”Energieautarkie“I Speicherung im simulierten MittelspannungsnetzI Ruckerzeugung (GT, GuD) im ubergeordneten Netz (nicht Teil der

Simulation)

11 / 20

Ergebnisse der Analyse

Jahresdauerkennlinie der P2G AnlageLandliches Netz Szenario 1

5. Combined Simulation 5.3. Analysis 50

0 3 6 9 120

20

40 PP2G,sc1

t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%

Figure 5.7.: Usage of the P2G plant during a year in the rural grid for scenarios 1 andcapping the installed capacity of P2G plant.

0 3 6 9 120

20

40

PP2G,sc2

t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%


0 3 6 9 120

2

4

6

8

10PP2G,sc1 = PP2G,sc2

t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 2500%

Figure 5.9.: Usage of the P2G plant during a year in the suburban grid for scenario 1and 2 and capping the installed capacity of P2G plant.

Landliches Netz Szenario 2


0 3 6 9 120

20

40 PP2G,sc1

t in monthsPP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%


0 3 6 9 120

20

40

PP2G,sc2

t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%


0 3 6 9 120

2

4

6

8


t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 2500%


Suburbanes Netz


0 3 6 9 120

20

40 PP2G,sc1

t in monthsPP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%


0 3 6 9 120

20

40

PP2G,sc2

t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G = 25%


0 3 6 9 120

2

4

6

8


t in months

PP

2Gin

MW

fP 2G = 100%fP 2G = 75%fP 2G = 50%fP 2G


= 25%

12 / 20


BetriebsmodiStandby mit PP2G,min = 15%Pn,P2G


00:00

04:00

08:00

12:00

16:00

20:00

00:00

0

5

10

15

20

t in daytime

Pin

MW

Eext

Ereg

00:00

04:00

08:00

12:00

16:00

20:00

00:00

0

5

10

15

20

t in daytime

Pin

MW

Estandby

Eusable

Figure 5.10.: Energies of the electrolyser working in the standby mode at April 20th inscenario 2 with Pn,P2G = 100%PP2G,sc2

5.3.2. The Standby ProblemThe standby power of an alkaline electrolyser has a linear dependence to it’s nominalpower [18],[32]1. In this work the assumption of the standby power is

Pstandby = 15%Pn,AEL . (5.24)

The problem of the standby power and therefore the resulting energy Estandby =E0 +

∫ t0 Pstandby(τ)dτ is only used to keep the electrolyser at the operating temper-

ature. The electrical energy is transferred into heat, but the produced hydrogen is notpure enough2 for a further use. For this reason an electrolyser typically operates abovethe standby power. The electrolyser needs to stay at least at Pstandby to prevent himof cooling down. If the electrolyser is cooled down, he is incapable to react on fastpower gradients, because according to Section 3.1.4 and he needs about 2h to reachthe operating temperature [18]3,[32],[45]. Another reason for the need of standby is thehigher life expectancy the electrolyser, because a periodical shut-down decreases theelectrolyser’s performance immense [4].

Summarised only a part of the energy consumed by the electrolyser is used to producehydrogen. The other part is used to keep the electrolyser at the standby mode. Figure5.10 shows the difference between the energies Eext, Ereg, Estandby and Eusable at a typicalday. The Energy Ereg is the amount of electrical energy produced by regenerative sources(WP and PV) in the grid, but not consumed by the load (private households and/orfarms), while the Eext is needed to keep the P2G plant on the operation temperature.

1In comparison to a PEM electrolyser the alkaline electrolyser is only useable above the standby power.PEM electrolysers are workable in 0 . . . 100% of the nominal power

2As a result of diffusion the porportion of O2 in the H2 stream is too high.3According to [18] the electrolyser usually remains well above the standby power and is only shut-downfor maintenance.

Teillast mit PP2G,min = 30%Pn,P2G5. Combined Simulation 5.3. Analysis 52

00:00

04:00

08:00

12:00

16:00

20:00

00:00

0

5

10

15

20

t in daytime

Pin

MW

Eext

Ereg

00:00

04:00

08:00

12:00

16:00

20:00

00:00

0

5

10

15

20

t in daytime

Pin

MW

Eusable

Figure 5.14.: Energies of the electrolyser working in the standby mode at April 20th inscenario 2.

100% 75% 50% 25%0

2,000

4,000

6,000

TFLin

h

Sc.1 (Standby)Sc.1 (Partial Load)Sc.2 (Standby)Sc.2 (Partial Load)Sc.3 (Standby)Sc.3 (Partial Load)

Figure 5.15.: Full load hours of the three scenarios at fP2G ∈ {100%, 75%, 50%25%}of the maximum appearing P2G power PP2G,scx of x ∈ {1,2,3} with thestandby and partial load mode.

A possibility to reduce the standby losses, is to use the electrolyser not less than inthe partial load of

Ppartial = 30%Pn,AEL (5.27)

instead of the standby power Pstandby [43]. Figure 5.14 shows the changed energies indifference to Figure 5.10. The advantage of this method is that Estandby is not used toheat the electrolyser, but produces hydrogen. Disadvantageous is the enhanced need ofexternal power and therefore external energy Eext. The usable energy Eusable results in

Euseable = Eext + Ereg (5.28)

raises significantly.

13 / 20


Wirkungsgrad der P2G Anlage

Wirkungsgrad im Standby Betrieb (helle Farbe) und Wirkungsgrad imTeillastbetrieb (dunkle Farbe) in Abhangigkeit zur bezogenen P2GNennleistung


0 3 6 9 120

1

2

3 ·106

t in months

VC

H4in

Nm

3

0 3 6 9 120

2

4

6

8 ·106

t in months

VC

H4in

Nm

3

Sc.1 100%Sc.1 25%Sc.2 100%Sc.2 25%

Figure 5.23.: Yearly produced VCH4 in the rural grid for scenarios 1 and 2, in standbymode (left figure) and minimal operating mode (right figure).

0 3 6 9 120

2

4

6 ·105

t in months

VC

H4in

Nm

3

0 3 6 9 120

0.5

1

1.5

2 ·106

t in months

VC

H4in

Nm

3

Sc.3 100%Sc.3 25%

Figure 5.24.: Yearly produced VCH4 in the suburban grid, in standby mode (left figure)and minimal operating mode (right figure).

100% 75% 50% 25%0

20

40

60

80

η P2G

in%

Rural, sc.1 (standby)Rural, sc.1 (min. operation)Rural, sc.2 (standby)Rural, sc.2 (min. operation)Suburban (standby)Suburban (min. operation)

Figure 5.25.: P2G efficiency ηP2G(1 a,fP2G) of the rural and suburban grid with fP2G ∈{100%, 75%, 50%, 25%} at standby and minimal operating mode.

14 / 20


Jahrliche Volllaststunden der P2G Anlagen

Volllaststunden im Standby Betrieb (helle Farbe) und Volllaststunden imTeillastbetrieb (dunkle Farbe) in Abhangigkeit zur bezogenen P2GNennleistung5. Combined Simulation 5.3. Analysis 57

100% 75% 50% 25%0

2,000

4,000

6,000

t FLin

h

Rural, sc.1 (standby)Rural, sc.1 (min. operation)Rural, sc.2 (standby)Rural, sc.2 (min. operation)Suburban (standby)Suburban (min. operation)

Figure 5.17.: Full load hours of the two scenarios at fP2G ∈ {100%, 75%, 50%25%} ofthe maximum appearing P2G power PP2G with the standby and minimaloperating mode.

Figures 5.18-5.20 show the higher Euse and Eext by comparing them to Figures 5.14-5.16. In some cases the external energy Eext is even bigger as the energy Ereg producedby renewable energy sources. For that the reason the demand on external energy increasesdramatically by operating the P2G plant in the minimal operating mode. The downside ofthe minimal operating mode is the higher external power consumption. By investigatingthe suburban grid it shows very well that a grid without WP needs a much higher externalenergy demand. The full load hours (5.30) are much higher with a higher Eusable, asshown in Figure 5.17.

15 / 20


Standby Batterie

Verhaltnis von aufzubringender Standbyenergie (helle Farbe) undspeicherbarer Energie (dunkle Farbe) in Abhangigkeit zur bezogenen P2GNennleistung5. Combined Simulation 5.3. Analysis 59

100% 75% 50% 25%0

20

40

Ein

GW

h

Rural, sc.1, Estandby

Rural, sc.1, Ecapped

Rural, sc.2, Estandby

Rural, sc.2, Ecapped

Suburban, Estandby

Suburban, Ecapped

Figure 5.21.: Standby energy Estandby and the capped regenerative surplus energy Ecappedat fP2G ∈ {100%, 75%, 50%25%}.

5.3.3. Standby BatteryThe idea of a standby battery is to reduce the power consumption of the high voltagegrid for the standby mode of a P2G plant, by installing a second storage system. It canbe done by storing the power peaks in batteries, instead of capping the power generation,as it is presented in 5.3.1. The capped regenerative energy

Ecapped(t,fP2G) = Ereg(t,100%)− Ereg(t,fP2G), fP2G = {75%, 50%, 25%} (5.33)

increases by a smaller installed P2G capacity. For that reason (with a battery) storableenergy increases to a maximum with fP2G = 25%. Figure 5.21 illustrates that the cappedrenewable energy for the rural and the suburban grid at fP2G = 25% is even bigger thanthe required standby energy. If fP2G = 100% the installation of a battery system isuseless because Ecapped = 0. The required battery size can be calculated as

Ebatt(t,fP2G) = Eext(t, fP2G)t

= Eext(1 a, fP2G)1 a , fP2G ∈ {100%, 75%, 50%, 25%}.

(5.34)

This energy Ebatt is an estimation of the minimum battery capacity required to reducethe external power consumption of a P2G plant for standby mode. For example during acloudy windless week, the battery capacity Ebatt is too small to guarantee the standbypower of the P2G plant. Figure 5.22 illustrates the minimum battery capacity dependingon the P2G plant’s nominal capacity.

The linear relation of Ebatt to Pn,P2G can be approximated by

Ebatt(Pn,P2G) = e0 + e1Pn,P2G. (5.35)

According to Figure 5.22 the straight line of scenario 1 in the suburban grid has the sametrend as the line of scenario 2 in the suburban grid. For that reason the approximation isdone for a battery in a grid with WP and PV (approximation 1) and in a grid with only

16 / 20

Schlussfolgerungen

Schlussfolgerungen - P2G Modell

I extrem dynamisch (15%→ 100%Pn in ≥ 6s) fallsI P2G hochgefahrenI keine Unterschreitung von Pstandby (AEL)

I alkalischen Elektrolyseur → PEM ElektrolyseurI besseres Teillastverhalten (0%→ 100%Pn)I kleine Leistungsgroßen bereits vorhanden

I AbwarmeI vom Elektrolyseur uninteressantI von der Methanisierung interessant

I kritische ParameterI Standby Leistung Pstandby

I interne Temperatur

17 / 20

Schlussfolgerungen

Schlussfolgerungen - Kombinierte Simulation

I optimal mit Kombination WP und PVI dynamische Abschaltung

I im Winter (nur PV)I in der Nacht (nur PV)I forecast (PV+WP)

I zusatzliche Batterie: Verwendung der cut-off EnergieI niedriger Power-to-Power Wirkungsgrad → anderweitige Nutzung von

H2, CH4 (z.B. Verkehr)

18 / 20

Schlussfolgerungen

Vielen Dank fur Ihre Aufmerksamkeit!

19 / 20

Quellen

1. ”Production of synthetic natural gas in a fluidized bed reactor”, Dissertation, Jan Kopyscinski2. ”Renewable Energies and Energy Efficiency”, Dissertation, Michael Sterner, 2009.3. ”Vorteile und Herausforderungen der Kopplung von Gas- und Stromnetzen”, Herbert Bauer, GASCADE Gastransport

GmbH, 2012.4. ”Modeling of advanced alkaline electrolyzers: a system simulation approach”, Ø. Ulleberg, International Journal of

Hydrogen Energy, vol. 28, no. 1, pp. 21-33, 2003.5. ”Design und Analyse: Systemgekoppeltes Verteilnetz der Zukunft”, O. Oberzaucher, Diplomarbeit, 2013.6. Abschatzung des Photovoltaik-Potentials auf Dachflachen in Deutschland”, M. Lodl, G. Kerber, R. Witzmann, C.

Hoffmann, and M. Metzger, 11. Symposium Energieinnovation, 2010.7. ”Das osterreichische Gasnetz”, C. Edler, TU Vienna, ESEA, 2013.

20 / 20

Anhang

Morgige Netzstruktur (Gas und Strom)2. Motivation 2.3. Implementation in Austria 6

70 bar

50 bar

16 bar

0.1 bar

Industry

Household (withPhotovoltaic)

Coal-Fired, Hydro-Electric,Gas-Fired Power Plants andCombined-Cycle Power Plants

Pump Storage,Wind Power

Photovoltaic,Wind Power

220 kV110 kV

20 kV10 kV

P2G

Gasimports

Pore Storage

Industry

Biogas0.4 kV

380 kV

Biomass

Figure 2.3.: Gas and electrical grid with typical levels and exemplary producers andconsumers. Picture based on [5]

processes. Besides the technical advantages P2G enables Austria to reduce it’s energydependency on countries like Russia. Germany already operates P2G pilot projects[2],[17].

2.3. Implementation in AustriaThe gas grid besides the electrical grid is the second largest power grid and the largestpower storage of Austria [15],[18],[16]. By comparing the power consumption of electricityin the year 2011 of 60.4TWh to the capacity of Austria’s usable gas storage capacityof 75TWh1 it is obvious that no other storage system is providing such a high storagecapability [12],[14]. Another possibility is to use the storage capabilities of the gas griditself, as an advantage to the electrical grid, the volume putted in has not to be consumedat the same time.

According to [8] Austria’s natural gas grid is differentiated in transmission and distri-bution levels. Gas transported in the transmission system does not directly supply theconsumers itself. The consumer’s gas supply is provided by systems in the distributionlevel which can be split up in

1calculated with the lower heating value of natural gas of [44]

11 / 20

Anhang

Elektrochemisches Modell

Zusammenhang Strom-Spannung

ucell = Urev + r1 + r2 T

Aicell + s log

(t1 + t2

T + t3T 2

Aicell + 1

)(1)

Produzierter Wasserstoff

nH2 = ηFNSNP

z F(2)

ηF =

(icellA

)2

f1 +(icellA

)2 f2 (3)

12 / 20

Anhang

Elektrochemisches Modell - Strom/Spannungskennline


0 50 100 150 200 250 300 350 400 4501.6

1.7

1.8

1.9

2

jcell in mA/cm2

ucellin

V

25 ◦C40 ◦C60 ◦C80 ◦C

Figure 3.5.: Cell voltage over the cell current density for different temperatures and asystem pressure p = 10bar

The parameters of the cell voltage are listed up in Table (A.1). The form of the voltageversus the current density

Jcell = Icell/A (3.27)

curve of (3.26) at the temperatures T ∈ {25 ◦C, 40 ◦C, 60 ◦C, 80 ◦C} is shown in Figure3.5. As can be seen the cell voltage is decreasing with increasing temperature at a stablevalue of current density. Usually a electrolyser stack consists of NS serial cells with theelectrolyser voltage

UAEL = Ustack = NS Ucell (3.28)

equal to the stack voltage. The electrolyser consists of NP parallel stacks, with theelectrolyser current

IAEL = NP Istack = NP icell (3.29)

Therefore the electrolyser consists of

Ncell = NS NP (3.30)

electrolyser cells as shown in Figure 3.6. This model is full scalable in the nominal powerof the electrolyser, by varying NS and NP . Usually NS stays constant and defines thelength of a 1MW stack. NP scale the plant up to n1MW and results in NP = n.

As shown in (3.26) the cell voltage of a electrolyser consists of the system pressure pand the system temperature T . The system pressure is defined and usually stays constantover the duration of a simulation. The system temperature will be calculated by the

13 / 20

Anhang

Elektrochemisches Modell - Wirkungsgrad

∂ηAEL∂Pn,AEL

< 0


00:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

00:00

300

600

900

1,200

1,500

t in daytime

Pin

W

Pin

0

0.2

0.4

0.6

0.8

1·107

Vin

Nm

3VH2

Figure 3.7.: Produced hydrogen in dependence to the input power Pin. The input equalsthe power of a typical PV production.

20 40 60 80 10060

70

80

90

100

Pin/Pn,AEL in %

η ein

%

Figure 3.8.: Efficiency of an alkaline electrolyser over the input power. The electrolysertemperature is constant at ϑ = 85 ◦C.

As a conclusion the output as a function of the electrochemical model fAEL,ec can bewritten by

yout,AEL,ec = fAEL,ec(xin,AEL,yout,AEL,th) =

NS Ucell

PinUAEL

ηFNS NP

z FVstd

UthUcell

. (3.45)

14 / 20

Anhang

Kombinierte SimulationAlgorithmus

5. Combined Simulation 5.1. Load Calculation 38

Start

Initialisation of theSincal database

Define thesimulation duration

Assignment in of load,PV and WP values

P2G plant sizeoptimationNext day

Read out of the values

Durationover

Visualisationof the results

Save the param-eter and values

End

yes

no

Figure 5.1.: Flow diagram of the Matlab® algorithm15 / 20

Anhang

Kombinierte SimulationOptimierung

5. Combined Simulation5.2. acronym]\glossaryentryP2G?\glossaryentryfieldP2G\glsnamefontP2GPower-to-

Gas\relax |setentrycounter[]page\glsnumberformat44P2GOptimation 44

Table 5.2.: Optimation algorithm to calculate the P2G plant size for one time step

Initialisation ∆0 ∈ (0,∆max] (Trust area)PP2G,0 ← Pprev or 0 (P2G start value)Pext,0 ← load-flow-calc (first load flow calculation via PSS®Sincal)k ← 1

repeat∆ = ∆0while Pext ≤ ∆ and Pext ≥ −∆ do

∆← 13∆

if Pext,k < 0 do feeding backPP2G,k ← PP2G,k−1 + ∆ increasing P2G powerif PP2G,k < 0 do only positive P2G power

PP2G,k ← 0else if PP2G,k < Pprev do reset

PP2G,k ← Pprevend if

else if Pext,k > 0 do feedingPP2G,k ← PP2G,k−1 −∆ decresing P2G powerif PP2G,k < 0 do only positive P2G power

PP2G,k ← 0else if PP2G,k > Pprev do reset

PP2G,k ← Pprevend if

end ifk ← k + 1 next iteration stepPprev ← PP2G,kPext,k = load-flow-calc load flow calculation via PSS®Sincal

until ‖ Pext ‖< ε or PP 2G = 0

16 / 20

Anhang

Energieautarkie

ηelE− = 36%E− ≡ E+

5. Combined Simulation 5.1. Load Calculation 40

00:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

00:00

−400

−200

0

200

t in daytime

Pin

W

PH0PP V,H0

00:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

00:00

−200

0

200 E+

E−

t in daytime

Pin

W

PH0 + PP V,H0

Figure 5.2.: The first plot shows the load power PH0 and the PV production PPV,H0 of aprivate household and the second plot shows PH0 +PPV,H0 with the energiesE+ and E−.

excluded. For this reason distribution grids are concentrically structured and (n-1)-safetyis ensured [19]. By simulating the rural grid the maximum PV power is

fPV,rural,H0 = fPV,rural,L0 = 87% (5.7)

of the installable PV power PPV,H0,rural and PPV,L0 of [28].In the suburban grid for Iline/Ith ≤ 75% the maximum power is fPV,suburban = 52% of

PPV,suburban. By the usage of a P2G plant in the middle of the grid according to Figure4.3 the maximum power can be lifted up to

fPV,suburban = 71% (5.8)

of PPV,suburban. The P2G plant reduced the power over the lines L01-L05 by consumingit, instead of transmitting it into transmission grid.

5.1.2. Energy AutarchyEnergy self-sufficient consumers produce exactly the same energy they are consuming. Itis only possible with the usage of a energy storage system. If energy is produced but not

17 / 20

Anhang

Standby Teillast5. Combined Simulation 5.3. Analysis 51

0

20

40

60

80

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Estandby

Eusable

0

20

40

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Estandby

Eusable

0

5

10

15

20

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Estandby

Eusable


0

50

100

150

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Eusable

0

20

40

60

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Eusable

0

10

20

30

100% 75% 50% 25%

Ein

GW

h

Ereg

Eext

Eusable

18 / 20

Anhang

Wochenlastgang100% und 25% P2G


0 1 2 3 4 5 6 70

10

20

t in days

PP

2Gin

MW

100%PP2G,sc2

25%PP2G,sc2

Figure 5.8.: Difference between the power consumption of the Pn,P2G = 100%PP2G,sc2and Pn,P2G = 25%PP2G,sc2 at the first week of April in scenario 2

75% PP2G,scx 50% PP2G,scx 25% PP2G,scx

0%

50%

100%Scenario 1Scenario 2Scenario 3

Figure 5.9.: Percentage of energy by a reduced P2G plant size to 75%, 50% and 25% ofthe maximum appearing P2G energy EP2G of the scenario x ∈ {1,2,3}.

Table 5.3.: Energy EP2G,scx(fP2G)/EP2G,scx(100%) of the four different P2G plant sizes,illustrated by Figure 5.9.

fP 2G 100% 75% 50% 25%

scenario 1 100% 99.4% 89.3% 68.1%

scenario 2 100% 99.5% 93.0% 70.1%

scenario 3 100% 98.0% 80.8% 49.5%

19 / 20

Anhang

Zyklen


100% 75% 50% 25%0

200

400

600

Ncycle

Rural grid, sc.1Rural grid, sc.2Suburban grid

Figure 5.26.: Cycles during a year in the rural and the suburban grid at fP2G ∈{100%, 75%, 50%25%}.

5.3.5. Cycle NumberThe dynamic and volatility of the P2G plant’s input power determines the cycles of aP2G plant. A cycle is similar to the charge cycle of a battery. In this work one cycle of aP2G plant is defined by raising the input power from

Pcycle,min = 30%Pn,P2G (5.40)

up to

Pcycle,max = 80%Pn,P2G. (5.41)

By decreasing the power down to Pcycle,min one cycle is finished21. The amount of cyclesis mode-independent, i.e. they stay constant if the P2G plant works in the standby modeor in minimal operating mode. Due to P ≤ Pcycle,min with P ∈ {Pstandby, Ppartial} andtherefore the lower bound is not exceeded variously.

Figure 5.26 shows the amount of cycles in both modes. The increasing number ofcycles by decreasing fP2G is well recognisable. For that reason the partial derivative ofthe number of cycles

∂Ncylce

∂fP2G< 0 (5.42)

in respect to fP2G is negative. This occurrence is explainable by the effect of reducingthe P2G plant’s installed capacity which causes a match of the capped regenerativepower peaks with one hysteresis cycle. Especially in the rural grid at scenario 2 andfP2G = 25% the cycle number is very high because in this scenario the WP is almost aspowerful as the PV.

The cycles are directly proportional to the life expectancy of a P2G plant. Unfortunatelyuntil now no scientific work discussing this relationship is published.21comparable to one hysteresis cycle

20 / 20

New Modellierung und kombinierte Simulation eines Power-to-Gas … · 2015. 7. 10. · Modellierung...

Documents

Transcript of New Modellierung und kombinierte Simulation eines Power-to-Gas … · 2015. 7. 10. · Modellierung...