Paul Scherrer Institut

Wir schaffen Wissen – heute für morgen

8. Juni 2015 PSI, 8. Juni 2015 PSI,

Paul Scherrer Institut

Can the decentralised CHP generation provide the flexibility required to integrate intermittent RES in the electricity system?

Evangelos Panos , Kannan Ramachandran

67th Semi-Annual ETSAP Workshop, Abou Dhabi

 Swiss energy system & Swiss energy strategy to 2050 objectives

 The concept of dispatchable biogenic CHPs (electricity-driven)

 Extensions on Swiss Times Electricity Model (STEM-E)

 Challenges in modelling

 Preliminary results from model testing

 Conclusions

Introduction

Net Imports Hydro Nuclear Thermal Renewables

Electricity sector Industry

Transport Residential Services -2

40 25

3 1

 Electricity production: 66 TWh

Swiss energy system in 2013

08.06.2015

PSI, Seite 3

 Final energy consumption: 896 PJ

 CO₂ emissions 41 Mtn:

0 50 100 150200 250 300

350 _Renewables

Heat Electricity Gas Oil Wastes Coal Wood

4 5

17 10

5

 Energy Strategy 2050 key objectives:

 Enhancement of energy efficiency

 Unlocking new RES (wind, solar, biomass)

 Withdrawal from nuclear energy

 Imports and fossil to meet residual electricity demand

 Extension of electricity grid

Dams south, Nuclear north

Swiss grid congested lines

08.06.2015

PSI, Seite 5

Source: Swissgrid

2/3 of the grid was built between 1950 and 60s with focus on

ensuring regional supplies from nearby plants

Gaseous fuel from waste biomass Wood

Air

Heat storage Energy conversion

with thermal

machine Electricity grid

 Combined Heat and Power plant

 Specifications for grid stabilisation:

 Fast response power generation

 Temporal independent production of electricity and heat

 A contractor (electricity utility) could operate a CHP swarm by remote:

 Dispatchable, scalable and decentralised power plant

 Balancing power is sold at high price levels on the market

The concept of biogenic CHP Swarm

Running project ETH/PSI sponsored by BFE and Swisselectric Research

 A biogenic flexible CHP can participate in the following markets:

a) Electricity supply, on-site or distributed

 competition with power plants

 competition with electricity grid price

b) Heat supply for space and water heating, on-site or distributed

 competition with boilers and heat pumps

 competition with district heating networks

c) Provision of grid balancing services at a grid distribution level:

 competition with on-site solutions e.g. batteries

 competition with services provided by pump hydro, PtG, etc.



However, the technology is resource-constrained:

 it uses biogas or upgraded biogas injected to gas grid

 needs access to gas pipelines

 competition with other biogas uses

Assessing the potential of CHP Swarms

08.06.2015

PSI, Seite 7

 To implement biogenic flexible CHPs in TIMES, to assess its

potential and to identify barriers and competitors we need in the model at least:

 Representation of decentralised power generation

 High time resolution to account for demand and resource fluctuations

 Introduction of dispatchability features (e.g. ramp-up constraints)

 Representation of heat supply and demand sectors

 Representation of electricity & heat storage technologies

 Representation of alternatives, such as power-to-gas pathways

 Representation of bio-methane production options

 Representation of demand for balancing services

Challenges in TIMES modelling

STEM-E:

 Swiss Electricity Model

 288 time slices:

4 seasons X 3 days X 24h

 Exogenous electricity

demand linked to economic activity

 Electricity load profiles

 Different types of power plants

 Resource potentials

Swiss Times Electricity Model

08.06.2015

PSI, Seite 9

STEM-HE:

 All STEM-E plus:

 Decentralised generation

 Heat demands & load profiles

 Heat supply options

 Storage for electricity & heat

 Power-to-gas / gas-to-power

 Upgraded biogas production

 Ramping constraints

 Balancing services

FROM STEM-E to STEM-HE

2012 2014

Representation of decentralised sector

Very High Voltage Grid Level 1 Nuclear

Hydro Dams Pump Storage Imports

Exports

Distributed Generation

Run-of-river hydro Gas Turbines CC Gas Turbines OC Geothermal

High Voltage Grid Level 2 Large Scale

Power Generation

Medium Voltage Grid Level 3 Wind Farms

Solar Parks Oil ICE

Waste Incineration

District Heating CHPs

Wastes, Biomass Oil

Gas Biogas H2

Low Voltage Distribution Grid

Large Industries &

Commercial CHP oil

CHP biomass CHP gas CHP wastes CHP H2 Solar PV Wind turbines

Commercial/

Residential Generation

CHP gas CHP biomass CHP H2 Solar PV Wind turbines

 4 different grid voltage levels are represented

 Allows for compensation of RES generation that is fed into grid

 Similar structure with TIMES-PET (and perhaps with other models)

 Not always straightforward assignment of power plants to each level

 To reduce complexity the focus is on heat that can be supplied by CHPs

 Space and water heating in buildings and commercial sectors

 Differentiation between different types of houses

 Two classes of heat in industrial sectors: <500

C, >500

C

Heat demand sector

08.06.2015

PSI, Seite 11

Residential Sector

Single Family Multi Family

Space Heating Existing buildings

Space Heating

New buildings

Space Heating Existing buildings

Space Heating

New buildings

Water Heating Industry

Heat

>500 ^oC (CHPs NO)

Services

Space Heating &

Water Heating

Heat Supply Technologies

(boilers, resistors, heat pumps, night storage devices, Decentralised CHP) Heat

<500 ^oC (CHPs YES )

 Statistical estimation of consumer behaviour from surveys ^[1]

 Data reconciliation to minimise the deviation between actual and calculated annual heat demand from the profiles:

Heat demands and load profiles

HDD Annual

Demand Outside

temperature Typical Daily Profile

% of seasonal demand

% of demand per typical day

% demand for 288 typical hours

Adjustment of electricity load curve

min� 𝑤𝑤_1,𝑡𝑡 𝐷𝐷_𝑡𝑡 − 𝐹𝐹_𝑡𝑡 ² +� 𝑤𝑤_2,𝑡𝑡 𝑥𝑥_{𝑡𝑡𝑡𝑡} − 𝑦𝑦_{𝑡𝑡𝑡𝑡} ²

𝑡𝑡𝑡𝑡 𝑡𝑡

𝐹𝐹_𝑡𝑡 𝑥𝑥₁… .𝑥𝑥_𝑛𝑛 = 0 𝐶𝐶𝑘𝑘 𝑥𝑥1… .𝑥𝑥𝑛𝑛 = 0

𝒘𝒘 weights, user defined 𝒚𝒚 Initial hourly profiles 𝒙𝒙 Adjusted hourly profiles

𝑭𝑭 Calculated annual heat demand 𝑫𝑫 Actual annual heat demand 𝑪𝑪 Other constraints that must hold

Single family houses Multi family houses

Commercial Industry

 Space heating: morning peak followed by a long day-time plateau and a smaller evening peak

 Water heating: sharp variations depending on use

Examples of obtained heat profiles

08.06.2015

PSI, Seite 13

Existing single family houses – Space heating in PJ

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

H01 H02 H03 H04 H05 H06 H07 H08 H09 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24

SPR-WK SUM-WK FAL-WK WIN-WK

Existing multi family houses – Space heating in PJ

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

H01 H02 H03 H04 H05 H06 H07 H08 H09 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

H01 H02 H03 H04 H05 H06 H07 H08 H09 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24

SPR-WK SUM-WK FAL-WK WIN-WK

All households– Water heating in PJ

Representation of heat systems

Heat supply option 1 Heat supply option 2 Heat supply option N

Heat storage

Heat System

𝑎𝑎

∙ 𝑋𝑋

+𝑏𝑏

∙ 𝑋𝑋

= 𝑐𝑐

∀𝑝𝑝 ∈ 𝑃𝑃

; 𝑝𝑝𝑝𝑝 ∈ 𝑃𝑃

; 𝑡𝑡 ∈ 𝑇𝑇 ; 𝑡𝑡𝑡𝑡_𝑝𝑝𝑝𝑝𝑎𝑎𝑝𝑝 ∈ 𝑇𝑇𝑇𝑇 ; 𝑎𝑎

, 𝑏𝑏

, 𝑐𝑐

, 𝑐𝑐

, 𝑐𝑐

𝑎𝑎

∙ 𝑋𝑋

+𝑏𝑏

∙ 𝑋𝑋

≤ 𝑐𝑐

𝑎𝑎

∙ 𝑋𝑋

+𝑏𝑏

∙ 𝑋𝑋

≥ 𝑐𝑐

Fixed capacity ratio: e.g. solar thermal and gas boiler

Upper and lower bounds on capacity ratios: e.g.

micro CHP and gas boiler

� 𝑋𝑋

𝑝𝑝∈𝑁𝑁_{𝑃𝑃𝑅𝑅𝑅𝑅}

≥ 𝑑𝑑𝑝𝑝𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑑

 Combinations of primary and secondary heating systems is possible via user constraints:

Heat demand

Avoiding technology mix in heat supply

08.06.2015

PSI, Seite 15

H01 H03 H05 H07 H09 H11 H13 H15 H17 H19 H21 H23

Wood boiler Solar thermal Gas boiler District heating Heat from CHP on-site Oil boiler

Heat Pump Electric boiler Coal boiler H01 H03 H05 H07 H09 H11 H13 H15 H17 H19 H21 H23

Wood boiler Solar thermal Gas boiler District heating Heat from CHP on-site Oil boiler

Heat Pump Electric boiler Coal boiler

 No individual technology

optimisation is applicable for heat supply systems:

a) Go for MIP by ensuring that only one heat system will supply a heat demand class ^[2] (not directly supported in TIMES) OR:

b) Introduce a utilisation curve for each heat supply system with NCAP_AF(UP) and ACT_UPS(FX) in accordance with the demand curve

 Primary reserves react to frequency deviations within 30 seconds

 Secondary reserves are activated just slightly after the primary reserves and maintain a balance between generation and demand within each balancing area (duration from 1 to 60 minutes)

 Tertiary reserves are called only after secondary control has been used for a certain duration, to free the secondary reserves for other purposes

Introducing balancing services

Secondary reserves

Secondary reserves Tertiary reserves

Source: Swissgrid

 Each power plant is producing 4 additional commodities related to the primary & secondary upward and downward reserves

 The set of the attributes of a power plant is augmented by:

 Its minimum stable operation

 The % of total capacity available for primary and secondary upward and downward reserves

 The ramp-up and ramp-down rates

 We can also provide the % of upward reserve met by online

plants to avoid unrealistically provisions of upward reserves from offline technologies

 The additional equations for balancing services:

 Can be introduced as UC (takes time to enter the constraints in EXCEL)

 Can be implemented as TIMES extension in GAMS (prone to errors)

Porting OSEMOSYS methodology

08.06.2015

PSI, Seite 17

 Each power plant eligible for participating in the balancing markets is divided into two parts

 A capacity transfer UC ensures that the capacity-related costs are paid only once:

 Demand for balancing services: 3 ∗ 𝜎𝜎

+ 𝜎𝜎

+ 𝐾𝐾

Porting OSEMOSYS methodology

𝑋𝑋

= 𝑋𝑋

Power Plant p (participation in the

electricity market) Power Plant pp (participation in the

balancing markets)

ELC Electricity demand Primary upward reserve demand

Primary downward reserve demand Secondary upward reserve demand

Secondary downward reserve demand

PU

PD SU SD

PU_DEM

PD_DEM SU_DEM SD_DEM

ELC_DEM

𝑋𝑋

= 𝑋𝑋

/(𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

∙ 𝑦𝑦𝑦𝑦𝑦𝑦𝑦𝑦

) :online capacity of process 𝑝𝑝 Downward reserve:

𝑋𝑋

≤ 𝑋𝑋

∙ 𝑑𝑑𝑎𝑎𝑥𝑥

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

: 𝑝𝑝 ∈ 𝑃𝑃𝐷𝐷, 𝑇𝑇𝐷𝐷

Upward reserve for a fast ramping plant: (𝐦𝐦𝒂𝒂𝒙𝒙

≥ 𝒔𝒔𝒔𝒔𝒂𝒂𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒑𝒑

) 𝑋𝑋

≤ 𝑋𝑋

∙ 𝑎𝑎𝑦𝑦

∙ 𝑑𝑑𝑎𝑎𝑥𝑥

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

: 𝑝𝑝 ∈ 𝑃𝑃𝑈𝑈, 𝑇𝑇𝑈𝑈

𝑋𝑋

+ 𝑋𝑋

≤ 𝑋𝑋

𝑋𝑋

≤ 𝑋𝑋

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

Upward reserve for a slow ramping plant: ( 𝒎𝒎𝒂𝒂𝒙𝒙

≤ 𝒔𝒔𝒔𝒔𝒂𝒂𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒑𝒑

) 𝑋𝑋

≤ 𝑋𝑋

∙ 𝑎𝑎𝑦𝑦

∙ 𝑑𝑑𝑎𝑎𝑥𝑥

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

: 𝑝𝑝 ∈ 𝑃𝑃𝑈𝑈, 𝑇𝑇𝑈𝑈

𝑋𝑋

+ 𝑋𝑋

+ 𝑋𝑋

∙ 𝑡𝑡𝑡𝑡𝑎𝑎𝑏𝑏𝑠𝑠𝑝𝑝𝑠𝑠𝑝𝑝

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡_𝑝𝑝 ≤ 𝑋𝑋

𝑋𝑋

+ 𝑋𝑋

+ 𝑋𝑋

≤ 𝑋𝑋

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

Key equations for balancing services

08.06.2015

PSI, Seite 19

Minimum online upward reserve ( 𝑝𝑝 ∈ 𝑃𝑃𝑈𝑈, 𝑇𝑇𝑈𝑈 )

∑ 𝑋𝑋

𝑝𝑝𝑝𝑝

≥ 𝐷𝐷𝐷𝐷𝑀𝑀

∙ 𝑑𝑑𝑚𝑚𝑑𝑑𝑠𝑠𝑑𝑑𝑠𝑠𝑚𝑚𝑑𝑑𝑝𝑝

All reserve from online plants when m𝑎𝑎𝑥𝑥

≤ 𝑡𝑡𝑡𝑡𝑎𝑎𝑏𝑏𝑠𝑠𝑝𝑝𝑠𝑠𝑝𝑝

: 𝑋𝑋

= 𝑋𝑋

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

Share of reserve from online plants when m𝑎𝑎𝑥𝑥

≥ 𝑡𝑡𝑡𝑡𝑎𝑎𝑏𝑏𝑠𝑠𝑝𝑝𝑠𝑠𝑝𝑝

: 𝑋𝑋

≥ 𝑋𝑋

∙ 𝑐𝑐𝑎𝑎𝑝𝑝𝑎𝑎𝑐𝑐𝑡𝑡

Upward reserve is limited by online capacity minus power output:

𝑋𝑋

− 𝑋𝑋

≥ 𝑋𝑋

+ 𝑋𝑋

Upward reserve by online plants limited by their max contribution to upward reserve:

𝑋𝑋

∙ 𝑑𝑑𝑎𝑎𝑥𝑥

≥ 𝑋𝑋

Key equations for balancing services

 The related to this work project is currently running and we are still integrating information from our partners regarding grid

constraints, balancing services, CHP technology characterisation and biomass resource potentials

 “Reference” scenario assumptions used to test the model:

 Based on the “POM” scenario of Swiss energy strategy 2050, implementing strong efficiency measures

 Fuel prices from IEA ETP 2014, translated to Swiss border pre-tax prices

 Nuclear phase out to be completed by 2034

 No CCS and no coal in electricity generation

 CO2 price rises to 58 CHF/t CO2 in 2050

 Solar potential: ~10 TWh, Wind potential: ~3 TWh, Hydro: ~40 TWh

Preliminary results from model testing

08.06.2015

PSI, Seite 21

Energy service demands in PJ

Forecast of demands

0 100 200 300 400 500 600

2010 2012 2015 2020 2030 2040 2050

Residential thermal

Residential electric specific

Services thermal

Services electric specific Industry thermal

Industry electric specific

(excl. transport)

Final energy consumption in PJ

08.06.2015

PSI, Seite 23

0 100 200 300 400 500 600 700

2010 2020 2030 2040 2050

Hydrogen Biogas Solar Wastes Wood Heat * Coal

Natural gas Heavy fuel oil Light fuel oil Electricity

Share of technologies in residential heat

0%

10%

20%

30%

40%

50%

60%

2010 2020 2030 2040 2050

Heat pumps Heat *

Gas boilers Oil boilers

Operational profiles of heat technologies

08.06.2015

PSI, Seite 25

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

H01 H03 H05 H07 H09 H11 H13 H15 H17 H19 H21 H23

Wood/Pelletts Gas

Heat radiators Oil

Heat pumps

Electric boilers/resistors Solar

Single family houses

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

H01 H03 H05 H07 H09 H11 H13 H15 H17 H19 H21 H23

Wood/Pelletts Gas

Heat radiators Oil

Heat pumps

Electric boilers/resistors Solar

Multi family houses

Storage technologies for heat in services

0 1'000 2'000 3'000 4'000 5'000 6'000

1 3 5 7 9 11 13 15 17 19 21 23 Total Heat

Demand -120

-100 -80 -60 -40 -20 0 20 40

1 3 5 7 9 11 13 15 17 19 21 23 Thermochemical

Storage of CHP heat Night boilers

Electricity generation sector

08.06.2015

PSI, Seite 27

0 2 4 6 8 10

2010 2020 2030 2040 2050

Solar Wind Wood &

Biogas

-10 0 10 20 30 40 50 60 70 80

2010 2012 2015 2020 2030 2040 2050

Solar Wind

Geothermal Wastes Wood Biogas Gas Nuclear Hydro

Pump storage Net Imports

Electricity production by fuel in TWh Electricity

production from

renewables in TWh

CHP Swarms (capacity & operation profile)

2020 2030 2040 2050

Capacity (MW) 109 185 124 56

Electricity production (GWh_e) 585 1200 522 415

% of total electricity production 1% 2% 1% 1%

% of decentralised thermal only electricity production 39% 80% 35% 28%

% of decentralised total electricity production 17% 16% 5% 3%

Heat production (GWh_th) 838 1719 748 595

0 20 40 60 80 100 120 140 160 180 200

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200

1 3 5 7 9 11 13 15 17 19 21 23

CHP Swarms MW Solar PV

(MW)

Solar PV CHP Swarms Winter working day

0 20 40 60 80 100 120 140 160 180 200

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200

1 3 5 7 9 11 13 15 17 19 21 23

CHP Swarms MW Solar PV

(MW) Summer Sunday

 CHP Swarms seems a promising technology for providing flexibility to the electric system

 Potential parameters affecting their uptake include:

 Developments in the large-scale generation

 Costs of bio-methane and access competition from other uses

 Feed-in tariffs for biomass

 Competition in heat supply from heat pumps

 Storage costs for storing excess heat from CHP Swarms

 Modelling challenges:

 No satisfactory solution for the technology mix effect in heat sectors

 Improvement of the dispatching of the power plant technologies:

crucial factor for the balancing services as well

Conclusions – Further Challenges

08.06.2015

PSI, Seite 29

References

[1] Main Sources used for obtaining the heat demand profiles:

 BFE, “Analyse des schweizerischen Energieverbrauchs 2000 - 2012 nach Verwendungszwecken“, 2013

 Rossi Alessandro, “Modelling and validation of heat sinks for combined heat and power simulation: Industry”, 2013

 Ayer Roman, “Modelling of heat sinks for combined heat and power simulation: Households”, 2013

 Federal office of Meteorology and Climatology MeteoSwiss

 Mark Hellwig "Entwicklung und Anwendung parametrisierter Standard-Lastprofile", 2003

 Ulrike Jordan, Klaus Vajen, “Realistic Domestic Hot-Water Profiles in Different Time Scales”, 2001

[2] Modelling heat systems:

 Merkel E., Fehrenbach D., McKenna R., Fichter W., Modelling decentralised heat supply: An application and methodological extension in TIMEs, Energy 73 (2014), 592-605

[3] Modelling balancing services:

 Welsch M., Howells M., et al., Supporting security and adequacy in future energy systems: the need to enhance long-term energy system models to better treat issues related to variability, Int. J. Energy Res. 39 (2015), 377-396

8. Juni 2015 PSI, 8. Juni 2015

PSI, Seite 31

Thank you for the attention ! Dr. Evangelos Panos

Energy Economics Group in the Laboratory of Energy Analysis

evangelos.panos@psi.ch