The Port of Rotterdam: the largest “capacitor” of the Netherlands

Delft, September 14th 2017 Rob Stikkelman r.m.stikkelman@tudelft.nl

2

E-Variability E-Flexibility (Demand, Supply) (Demand, Supply)

The Harbour Industrial Cluster (HIC) of Rotterdam: • represents a huge power system, • potentially offers significant E-flexibility However,….

The “Flexible Energy in Rotterdam” project

3

There was a lack of data and knowledge: • Which flexibility options can we create within

individual companies and clusters? • What is the effect of integrated processes? • How much variability do we observe in the electricity

markets? • How do we harvest this variability? • To which extent is the electric power grid a limiting

factor? • Etc.

The “Flexible Energy in Rotterdam” project

4

Project team

The “Flexible Energy in Rotterdam” project

Markets

The Harbor Industrial Cluster as “Capacitor"

HIC Processes

net

wo

rk

€

up down

Natural gas

Electricity products

5

6

Markets: Average prices Gas and Power [Euro/GJ]

Gem. Transactieprijzen Aardgas (blauw) en Elektriciteit, grootverbruik ex BTW incl. Taks Bron CBS

7

Minimum: 1,7 Average: 40,1 Maximum: 99,8

Market: Day Ahead prices 2015 [pte, Euro/MWh]

34.00032.00030.00028.00026.00024.00022.00020.00018.00016.00014.00012.00010.0008.0006.0004.0002.0000

100

95

90

85

80

75

70

65

60

55

50

45

40

35

30

25

20

15

10

5

8

E

Market: Prices RRV trade 2015 [pte,Euro/MWh]

34.00032.00030.00028.00026.00024.00022.00020.00018.00016.00014.00012.00010.0008.0006.0004.0002.0000

700

600

500

400

300

200

100

0

-100

-200

-300

-400

Up Down Minimum: 0 -429 Average: 79 7 Maximum: 738 78

9

Market: Prices imbalance trade 2015 [Euro/MWh]

<price> Euro/MWh

<Volume> MWh/pte

UP 79 15,1

DOWN 7 -13,9

10

E

Networks: 150 and 380 kV net

11

E

Selected processes and products

• Electrolysis of brine, Chlorine, Cl2

• Production of liquid Carbon dioxide, CO2

• Air separation, Liquid Oxygen O2

• Charging Batteries of Automated Guided Vehicle (AGV)

• Milling Cement

• Cooling warehouses

• Production of steam

• Heating of stored bunker oil

12

Example of a flexible process: steam production

Electricity

Steam Demi water

Electric boiler

Gas boiler Gas

Emissions

€

13

Modelling the system: Linny-R by Pieter Bots

• easy-to-use software tool for industrial process optimization (LP)

• allows to model a production system as (clusters of) processes that have products as inputs and outputs.

• calculates at what level processes should be scheduled to produce the desired quantities of end products while maximizing profit (or minimizing cost).

• Allows time series

• Uses the lp_solve library version 5.5.2.0 developed by Michel Berkelaar, Kjell Eikland,

and Peter Notebaert, • Uses the API for Delphi developed by Henri Gourvest

14

Modelling the system: Linny-R

15

Network: A (too) Simple model OUTCOMES OF EXPERIMENT "Base run" -- Success

0

0,001

BLStroom[MWh]

0

0

E invoeruit markt

[MWh]

1

51,70,0518

1 0

1

0

1

0

10

1

51,70

51,7[0 ... 70]

0,001

BLbasislast

0[0 ... 20]

0,01

BL hogelast

0[0 ... 10]

0,1

BL Extremelast

Maasvlakte Geervliet Europoort Theemsweg Merseyweg Botlek Oudeland

16

Modelling the system: The production of chlorine

0

25,0

Liquifiedchlorine

[ton]

0

7,27

NaOH32%[ton]

0,54833

H2[ton]

39,6

11,4

NaOH50%[ton]

-29,70

NaCl[ton]

-28,50

DemiWater[ton]0

10,0

power[MWh]

-61,9

0

low-pressuresteam

[MWh,th]

1095

2,96

<1600

>500

Cl2storage

[ton]

16

25,0

=16

Cl2delivered

[ton]

-5

-150 -150

imbalancepower[MWh]

-40

30 30

day-aheadpower[MWh]

1 18450

3,4461,91547

0,03

0,5413,5

161,9450

0,6439,6288

1,6529,70

2,82

50,80

2,5

45450

1

61,90

12

50,0

1 250,0

10

10

116400

1 16400

0,3622,3162

1

401200

1

401200

1

5-750

1

5-7501

0

1

0

18[9 ... 18]

25,0

Electrolysis

61,9[0 ->

7,27

Evaporator

2[0 ->

25,0

fill

0[0 ->

2,92

empty

16[0 ->

25,0

Offtake

5[5]

-150

buy onimbalance

market

40[40]

30

buy on day-ahead market

0[0]

10,0

sell onimbalance

market

Joint effort with companies to:

• build trust and mutual understanding

• create curiosity

17

Modelling the system: standard storage process

OUTCOMES OF EXPERIMENT "Base run" -- Success

2,5

17

=2,5

productSTO1[MWh]

171

0

<240

>120

opslagSTO1[MWh]

0

0,001

THStroom[MWh]

-65,2

34

<0

E day-ahead[MWh]

0

0

=0

Eafregelen

[MWh]27,9

34

<37,7

>0

Eopregelen

[MWh]0

34

inkooptotaal STO1

[MWh]

11,250

1 1,2542,5

11,250,0013

12,585

10

1

2,585

1 0 11,2542,5

1 1,2542,5

1

1,2542,5

1 0

1

01

1,250

0[0 ->

17

vullenSTO1

1,25[0 ->

0

legenSTO1

1,25[1,25 ... 3,75]

34,0

producerenSTO1

2,5[2,5]

34

DA inkoopSTO1

0[0 ->

0

E inkoopSTO1

1,25[0 ->

34

E verkoopSTO1

STO STO STO STO STO

18

Modelling the system: standard substitution process

SUB SUB SUB SUB SUB

19

Results: Single storage process and imbalance prices

Effect of storage size and future knowledge of prices

20

The port model OUTCOMES OF EXPERIMENT "Base run" -- Success

2,5

17,5

=2,5

productSTO1[MWh]

0

0

=0

Eafregelen

[MWh]-65,2

35

<0

E day-ahead[MWh]

-139

25

<0Gas

[MWh] 0

0

E invoeruit markt

[MWh]

0

0,001

EPStroom[MWh]

0

0,001

BLStroom[MWh]

0

0,001

MVStroom[MWh]

0

0

GVStroom[MWh]

0

0,001

OLStroom[MWh]

0

0,001

THStroom[MWh]

0

0

MEStroom[MWh]

45

19,4

=45

ProductSTO2[MWh]

0,25

3,5

=0,25

ProductSTO3[MWh]

9,25

23,5

=9,25

ProductSTO4[MWh]

0,125

0

=0,125

ProductSTO5[MWh]

0,75

31,4

=0,75

ProductSTO6[MWh]

1,5

31,4

=1,5

ProductSUB1[MWh]

30

31,4

=30

ProductSUB2[MWh]

45

31,4

=45

ProductSUB3[MWh]

22,5

31,4

=22,5

ProductSUB4[MWh]

15

31,4

=15

ProductSUB5[MWh]

2,5

9,35

=2,5

ProductSOP1[MWh]

9,5

9,4

=9,5

ProductSOP2[MWh]

3

9,21

=3

ProductSOP3[MWh]

24,7

35

<28,1

>0

Eopregelen

[MWh]

1

1,2543,8

1

0

1

0

1

2,20

1

2,20,0022

1

0

1

0

1

27,50,0276

1

0

1

0

1

0

1

0

1

27,50

1

0,69750,0007

1

0

1

0

1

0

1

0

1

0,69750

1

0

1

0

1

0

1

0

1

0

1

0

1

0,353,50E-04

1

0

1

0

1

0

1

0

1

0,350

1

9,70,0097

1

0

1

0

1

0

1

0

1

9,70

1

0

1

0

1

0

1

0

1

0

1

0

1

1,250,0013

1

0

1

25875

1

250,025

1

0

1

0,0250,875

1

0,0252,50E-05

1

0

1

6,2217

1

6,20,0062

1

0

1

0

1

0

1

0

1

0,672523,5

1

0,67250,0007

1

1,547,2

1

0,0757,50E-05

1

1,7844,5

1

30943

1

1,50,0015

1

35,6891

1

451415

1

2,250,0023

1

53,41336

1

22,5707

1

1,130,0011

1

26,7668

1

15472

1

0,750,0008

1

17,8445

1

0

1

0,71523,4

1

0,2752,75E-04

1

0,5513,8

1

0

1

2,7389,3

1

1,050,0011

1

2,152,5

1

0

1

0,84527,6

1

0,3253,25E-041

0,6516,2

140,50

1

451575

1

0

1

20700

1

0,1254,38

1

0

1

0,1254,38

1

2,587,5

1

0

1

1,2543,8

1

9,25324

1

0

1

3,05107

1

0,258,75

1

0

1

0,2257,87

1

0,7526,3

1

0

1

0,07752,71

1

2,2578,8

1

0

1

0

1

1,0536,7

1

0

1

0

1

0,0752,63

1

0

1

0

1

1,552,5

1

0

1

0

1

1,1339,4

1

0

1

0

1

0,7526,3

1

0

1

0

1

0,2759,63

1

0

1

0

1

0,32511,4

1

0

1

0

1

0

1

1,250

1

0,2250

1

0,07757,75E-05

1

2,160,0022

1

200

1

6,770,0068

1

3,050

1

0,1250

1

1,780,0018

1

0

1

0

40,5[0 ->

0

LeveringElektriciteit

0[0 ->

100

Andereafregelaars

0[0 ->

1000

Andereopregelaars

STO1

Europoort 150 kV200 MW

Botlek 150 kV400 MW

Maasvlakte 150 kV200 MW

Geervliet 150 kV200 MW

Oudeland 150 kV200 MW

Theemsweg 150 kV200MW

Merseyweg 150 kV200 MW

STO2STO3 STO4 STO5STO6 SUB1SUB2 SUB3SUB4 SUB5 SOP1SOP2SOP3

Natural gas Electricity production

Up Down

Day-Ahead market

21

Experiments for 2015

1. Reference case

2. No balancing

3. Zero flexibility

4. No limits

– Day- Ahead and RRV prices

– Volumes Up and Down RRV per pte

– Gas price is 25 €/MWh

– As reference case, but volumes Up and Down are zero

– As No balancing, but no gas tipping point

– As reference cases, but infinite volumes for Up and Down

22

Results: Reference case

Mean

St.dev.

Min.

Max.

Sum

[MWh/pte]

[GWh/year]

Down BM

-6,1

12,5

-122

0

-210

Up BM

1,0

3,4

0

66

35

Day-ahead

-68,8

22,5

-210

-65

-2377

E-load 73,8 24,1 37,3 203,9

2550 G-Load 141,3 24,8 14,7 152,1

4883

Flex load HIC 2015 4*(1,0 − -6,1)= 28,4 MW

23

Results: No Limits

Mean

St.dev.

Min.

Max.

Sum

[MWh/pte]

[GWh/year]

Down BM

-41,3

56,3

-139

0

-1428

Up BM

3,4

13,9

0

173

116

Day-ahead

-68,8

22,5

-210

-65

-2377

E-load 107 59,3 37,3 204

2550

Upper limit average Flex load HIC

4*(3,4 − -41,3) = 179 MW

Max transport capacity 4*173= 692 MW

Network OK?

24

Results: Value of Flexibility

Sum

Energy costs

[M€/year]

Zero flexibility 221,7

No balancing 219,9

Referentie casus

205,8

No limits

157,6

Value flex HIC in 2015 221,7 − 205,8= 15,9 M€ Market volume limits

Maximum value flex 221,7 − 157,6= 64,1 M€

Technique limits

25

Results: Network performance at No Limits

0

5000

10000

15000

20000

25000

30000

35000

40000

0-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100

Ouweland

Theemsweg

Europoort

Maasvlakte

Botlek

Percentage of maximal transport capacity [%]

Nu

mb

er o

f p

te

Bottleneck?

26

Closure

Model results for 2015 case:

• The port of Rotterdam is able to offer on average 179 MW of flexibility

• The RRV market offered on average 53,2 MW of variability

• Due to market limitations only 28 MW on average was used with a total value of 15,9 million Euro

• Value of the RRV market in 2015 was 32,4 million Euro

27

Closure

The project was a success:

• Project members were always on board

• We learned a lot

• Follow-up projects are developed

Simple LP model has been made:

• New insights were generated

• The model can easily be adjusted for other industrial area

• Too simple?