Renewable energies play an important role in climate change mitigation (International Energy Agency 2017). When introduced massively, integration of renewables on a different scale from stand-alone buildings to cities and nations or regions will be necessary. The challenges ahead with the decarbonization of the energy systems are huge as major part of the carbon emissions needs to be eliminated by the middle of this century. For example, the European Union has planned to reduce its emissions by about 80% to 95% and to reach a 55% share of renewable energy in final energy consumption by 2050 (Oettinger 2012). In addition to carbon emission reductions, access, security, reliability, and affordability need to be addressed in the energy transition (Dastur 2009).
While the clean energy transition is often dominated by variable renewable electricity sources such as solar PV and wind power for power production, the future focus needs also to better include the heating sector, as most of the final energy use is in the form of thermal energy. For example, in Finland over 80% of the final energy use of households is for heating and domestic hot water (Statistics Finland 2018). In Helsinki, which is the capital of Finland, heat constitutes 60% of whole energy production (Lund, Mikkola and Ypyä 2015). Therefore, the role of the built environment in the energy transition will also be important to be considered; measures of interest include building energy efficiency and on-site renewable energy generation (Lund 2012, Cao et al. 2013, Kayo et al. 2014, WBCSD 2009).
Another interesting development is the need to increase energy system flexibility when the share of variable renewable electricity (VRE) increases, which may include measures such as curtailment of VRE for better energy system management, but also to supply heat through power-to-heat (P2H) conversion e.g. when surplus of VRE were available (Lund et al. 2015, IEA 2008, IEA 2009). One may also foresee ‘oversizing’ VRE for the power sector needs to provide high coverage of power demand with VRE during less favorable condition, but then use the surplus or curtailed power at favorable conditions for heating purposes. This would mean increased electrification of the heating sector, and if linked to e.g. heat pumps, providing large amounts of renewable-derived heat to end-users. This kind of sectoral coupling of VRE would increase the energy system flexibility, but reduce emissions in the heating sector as well. Efficient VRE integration has been studied from different aspects (Bessa 2014, Brown 2013, Galus 2013, Sharif 2014, Gao 2008). Other measures such as energy storage and demand response could also be used in this context (Arabzadeh et al. 2018, Lund 2018).
The novelity of the current study is examining of the VRE coupling with the heating sector using different curtailment strategies not yet reported in the literature, in particular for both the national and city level production systems which are included in this study. For the national level analysis, we use Finland as the case to see how the whole heat production portfolio would change with adding large wind power schemes with sector-coupling to heat through curtailment and P2H. For the sub-national level, we use Helsinki city as the case. In this case, considering the energy system in more detail than the national scale will be important, enabling then to assess how the different VRE curtailment schemes may affect the power plant uses. In particular, existing CHP-based energy production system, which is typical for northern cities such as Helsinki, is analyzed, including the running costs and emissions on a yearly scale. Also, the effects of different energy system limitations and building energy efficiency schemes are analyzed.
The methodology employed in the study is based on a techno-economic optimization of the energy system, i.e. finding the system configuration which yields the minimum cost under different boundary conditions and limitations. We consider both national (Finland) and sub-national (Helsinki city) levels of the energy production system, for which different optimization techniques will be applied. As clean energy transitions often involve large renewable electricity usage, we take here as the starting point a large wind power scheme. This will be accompanied by different curtailment strategies and power-to-heat conversion to produce carbon-free heat, which is of importance in northern latitudes. Curtailment may also be justified by energy system limitations and power market conditions, thus providing better energy system management.
The national-level analysis is done with a macro-scale energy system model (Pilpola and Lund 2018). The model includes all aspects of an energy system, including electricity, heat, and fuels, thus covering all energy sectors. The model employs a 1-hour step for electricity and heat, while fuel demands are considered on an annual scale. The model seeks for a cost-optimal solution of the energy system while keeping the supply-demand at balance. The optimization problem is defined by Equation (1):
The variables in the optimization are the amounts of the fossil primary energy sources and the amounts of conventional conversion, i.e. combined heat and power (CHP), separate production, and heat pumps (HP). The main optimization outputs are the primary energy composition and power and heat production, while the main inputs are historical consumption and temporal data, cost assumptions, and system constraints. The model uses year 2013 as the reference year for input data. A more detailed description of the input data can be found in (Pilpola and Lund 2018). The reference case for the heating sector of interest is shown in Table 1.
|CHP-District heating||CHP-Industrial||Separate thermal production||Residential fuel boilers||Residential electric boilers||Heat pumps||Total|
The hourly distribution of the conventional production, such as CHP, is based on historical production data (2013) to mimic the hourly distribution, but their values are scaled to the yearly amounts of energy (PJ, TWh) resulting from the optimization. The operation of the power-to-heat conversion is rule-based, e.g. linked to VRE levels. The level of industrial CHP and residential heat production (fuel and electric boilers), which accounted for 43% of the heat demand in 2013, are assumed non-variable. The model employs three methods for wind power curtailment (described separately). The model also assumes 60 GWh thermal storage available through the Finnish district heating networks filled with water.
The objective of the sub-national (city) model is to reduce the marginal costs of the local energy production system while introducing variable renewable electricity. The model is based on an optimizer core frame, minimizing a cost function, which is the sum of the production costs, balancing costs, storage costs, and revenues from electricity sales to the Nordic power exchange. Equation (2) shows the cost function:
The details of the model are described in (Mikkola and Lund 2016). The model includes 3 main sets of constraints: the technical properties of the power plants, technical properties of the balancing methods, and the balance of energy demand and supply. Therefore, the frame is appropriate to model the energy production system. We also incorporated several modifications to the original model to enable a heating sector analysis as follows: Firstly, a limitation for heat storage charging and discharging was defined. Secondly, to account for possible heat source restrictions in case of large heat pump schemes with P2H conversion, we defined a dynamic COP, which drops when entering the peak heat demand period (HP has a COP = 3 whenever the heat demand is <50% of the peak heat demand; COP = 2 for heat production between 50% and 70% of the peak demand; COP = 1, i.e. HP is replaced with an electric boiler, when heat demand is >70% of the peak demand). The annual heat demand in Helsinki (60°N) is 6.6 TWh and the electricity use is 4.4 TWh. Two energy efficiency strategies in buildings will be considered leading to 8% and 16% lower heating demand, respectively. The heat demand profiles will be modified accordingly in these cases by using a scaled U-value for the temperature-dependent part of the heat load. The details of the present energy production plants in Helsinki are shown in Table 2 (Mikkola and Lund 2016). In addition, the district heating system has 5 GWh of storage capacity through its piping network, which can be employed for storing curtailed power in from of thermal energy.
|CHP gas||CHP coal||CHP coal||Boiler gas||Boiler Oil||Boiler Coal||HP|
Three different methods for wind energy curtailment will be employed, illustrated in Figure 1. “Peak-shaving” curtailment is based on shaving a fixed percentage of the wind power peaks. In “Wind-following”, wind power is curtailed with a constant fixed percentage. In “Load-following”, wind power is curtailed above a fixed level (percentage) of the electrical load. The curtailed wind power is used with a power-to-heat (P2H) strategy, i.e. all curtailed wind power will be directed to heat pumps. As wind power is a renewable and zero-marginal-cost resource, we argue that the heat produced by this “forced” heat pump operation offers a renewable and low-cost, albeit non-dispatchable, source of heat, which in turn may have the possibility to decrease the marginal costs of energy production. In the sub-national level, the forced implementation of the HP, limits the available HP capacity for the dispatch optimization.
In this section, we analyze the effects on the heat production system when applying large wind power schemes and curtailment with the P2H conversion. On the national level, the effect of each curtailment method is discussed in term of system operation and energy conversion. On the sub-national level, the focus is on the heating demand profile modification through building energy efficiency measures in addition to heat resource limitations for heat pumps.
In the national case, the three wind power curtailment strategies described previously resulted in very different levels of curtailed wind power shown in Table 3. Two levels of wind power use were investigated: 20 TWh and 40 TWh, corresponding to 24% and 48% of the yearly electricity use of Finland. The “Peak-shaving” strategy yielded the least heat, even at a high curtailment level. The “Load-following” did not produce any heat with 20 TWh of wind, but clearly more with 40 TWh than the “Peak-shaving” strategy. “Wind-following” gave the highest level of curtailment: With 30% curtailment and 40 TWh of wind it resulted in 36 TWh of heat production, which would account for 40% of the total annual heat demand. Note that the curtailed wind power was used in heat pumps (COP = 3). Figure 2 shows the change in the heat production mix caused by the different curtailment and P2H schemes. The reference case is Finland in 2013, based on historical data. Most cases exhibited increased CHP production (particularly coal-CHP if the CO2 and fuel cost remain low) and decreased separate heat production. As decrease of boiler use was also observed in the cases without curtailment and P2H, it can be stated that this effect was caused by the overall cost optimization, suggesting that with increasing wind power, replacing separate heat boilers with CHP may be cost-effective even without any additional P2H measures. However, including curtailment+P2H would emphasize the role of the heat pumps, i.e. the HP heat production increased in all cases with wind power curtailment. The increase in heat pump production was partly caused by the “forced” wind power curtailment strategy, but also partly by the overall cost optimization. Some CHP would, however, remain despite the increased HP production due to the economic reasons: it may be cost-effective to produce CHP-based electricity to be exported to international markets, even if that may simultaneously lead to oversupply of heat.
|Wind power before curtailment (TWh)||Curtailment rate (%)||Curtailment method|
|Peak-shaving (TWh)||Wind-following (TWh)||Load-following (TWh)|
Figure 2 also illustrates that in cases of a higher amount of wind-power driven (“forced”) heat pump operation, the net change of heat production is positive. As the overall heat demand stayed constant in all cases, this positive net change suggests that part of the additional heat production was wasted. In the “Wind-following” curtailment case, heat production was up to 20 TWh higher than the demand. A part of the excess heat production may appear in the losses of thermal storage, but clearly forced, non-dispatchable heat pump operation was not always able to match the heat demand, which leads to oversupply of heat. Some of the surplus heat could, however, disappear if the COP of the heat pumps were lower than anticipated, which could occur in particular during the peak heating season. Increasing the thermal storage capacity could also help to mitigate the heat surplus.
For the sub-national level, the effect of heating demand improvement (energy efficiency), wind curtailment, and heat source limitations of heat pumps were investigated through four scenarios:
In all scenarios, building energy efficiency improvements are applied: present building U-value is improved by 10% and 20% leading to 8% and 16% lower heat demand, respectively. The building energy efficiency improvement reflects the potential for energy efficiency in the old building stock, typical to Helsinki. The necessary measures could typically involve better thermal insulation, improved heating control, and/or better windows in the building stock. Moreover, 750 MW and 1500 MW of wind power was introduced to the energy production system, representing 43% and 86% of the yearly electricity demand in Helsinki, respectively. The 1500 MW heat pump scheme covers almost 50% of the peak heat demand, the rest of the peak being supplied through boilers. For comparison, the yearly fuel costs and CO2 emissions are compared with the present system with no measures, for which the system parameters are given in Table 2. The results for Scenario 1 are shown in Table 4. In this section, in all figures, the term “norm” means heat production with heat pump by electrical power and “curt” means the produced heat with heat pump by curtailed wind power.
|Case||Heat demand reduction (%)||Wind (MW)||Fuel cost (M€/yr)||Emissions (ktCO2/yr)|
Wind power alone would mainly reduce the gas use, but not coal, because of the need to meet the heat demand (coal-CHP produces more heat than gas-CHP). Also, the price difference between coal (€43/MWh) and gas (€56/MWh) would slightly favor coal. No emission cost was applied, but e.g. a €15/MWh CO2 cost would not change the results due to these system limitations. Wind power would therefore mainly reduce the fuel costs, but marginally the emissions in the Helsinki case with the assumptions used.
Building energy efficiency measures in cases 4 and 6 gave the best results with cost reductions of 19% and 26% and emission decreases of 7% and 12%, respectively. Detailed heat and electricity production are shown in Figure 3, which clearly demonstrates that most of the wind power would go to export to the Nordic power exchange. Note that Helsinki presently produces more electricity than it needs, i.e. some 40% is sold to the market (export).
In Scenario 2, 1500 MW of heat pumps is added to Scenario 1. Neither curtailment nor limitation of heat sources for heat pumps is included here. Table 5 presents the key results.
|Case||Heat demand reduction (%)||Wind (MW)||Fuel cost (M€/yr)||Emissions (ktCO2/yr)|
Adding a large heat pump scheme would significantly reduce the yearly fuel costs and emissions, whereas adding building energy efficiency measures on the top of that had less effect. In the best case 6, the fuel costs and emission dropped by 55% and 58%, respectively. Figure 4 demonstrates details of the energy production for the Scenario 2. There are slight differences only between cases 4 and 6, which indicates that reducing the heat demand from 8% to 16% does not have any major effect on the operation of the energy production system. Integration of the wind power reduces the gas electricity production by about 42% and the coal electricity production about 70%, respectively. The gas heat production is reduced by 31% and coal heat production is cut 77%. The HP heat production increases by 70%. The heat storage use increased by 70%, i.e. the number of charging-discharging cycles of the 5 GWh capacity in the DH network increased substantially.
In Scenario 3, wind energy curtailment was introduced to Scenario 2. The resulting fuel cost and emissions are in Table 6.
|Case||Heat demand reduction (%)||Curtailment method||Curtailment (%)||Fuel cost (M€/yr)||Emissions (ktCO2/yr)|
Figure 5 demonstrates the details of the effects on the energy production system. The largest reductions in the fuel cost and emissions were for case 6, 66% and 68%, respectively, when the “Load-following” curtailment method was used, though the difference to “Wind-following” curtailment was small. The heat production by coal was dramatically cut to less than 10% of the yearly heat production (–84%); the gas use dropped below 30% (–47%), respectively. The reduced coal and gas use lead to significantly reduced emissions (66-68% from the present). Heat pumps are the dominating heat production method in cases 4 and 6, which leads to a slight increase in electricity imports.
In Scenario 4, we added a heat source limitation to heat pumps by using a dynamic COP explained earlier to reflect the changes in the heat source availability for the heat pumps. An electric boiler with a maximum output of 1500 MW (equal to COP = 1) was therefore added to the energy production system. Table 7 presents the results for Scenario 4.
|Case||Heat demand reduction (%)||Curtailment method||Fuel cost M€/yr||Emissions (ktCO2/yr)|
The reference case in this scenario is a modified system since the nominal maximum HP output is set as 1500 MW. For each curtailment method, the lowest fuel cost and emission were achieved with the highest heat demand improvement of 16% (case 3, 6 and 9). The details of the energy production for Scenario 4 are shown in Figure 6. The base system in this scenario is the modification of the present system (1500 MW heat pumps), but electric boilers are used in cases 6 and 9 due to heat source limitations. In case 9 with the “Load-following” curtailment scheme, there is no heat storage use at all. It seems that increasing the wind energy share would reduce the contribution of the heat storage in this case. Both the share of gas and coal in heat production is halved compared to the base system; the share of gas comes down to 25% and coal to 5% of the yearly heat consumption. In power production, we see the similar trend: the share of gas power production is halved compared to the base system to 43% of power consumption and coal 5%, respectively. Due to the heat resource limitation, a part of the heat consumption from curtailed wind power is supplied by the electric boiler. For the “Load-following” curtailment (case 9), the electric boiler contribution is 18%. In Figure 6, the term “curt” and “norm” for the electrical boiler deal with heat production by electrical power and produced heat with curtailed wind power.
In this paper, we have studied the sectoral coupling of variable renewable energy (VRE) to the heating sector on a national (Finland) and city (Helsinki) level. For VRE, we used a large wind power scheme to which different curtailment & P2H strategies, building energy efficiency measures, and heat pump schemes were added. The effects to the existing energy system were analysed with dynamic energy system models. On the city level, four basic scenarios with a total of 27 cases were analyzed to reveal the system effects from the above strategies, while on the national level 14 cases were analyzed. The main findings of the analyses are the following:
The heat pumps played a major role in the sectoral coupling of VRE to the heating sector. On a national level the heat pumps were assumed to work at a constant COP, but on the city level we employed a dynamic COP which considers heat source limitations. The assumption of a constant COP may lead to too optimistic values of heat production, but as this also caused oversupply of heat which couldn’t be utilized, the error to the energy balance is probably minor.
This work was supported by the Nordic Energy Research [grant number 76084] and the Academy of Finland [grant number 285353].
The authors have no competing interests to declare.
Arabzadeh, V, Alimohammadisagvand, B, Jokisalo, J and Siren, K. 2018. A novel cost-optimizing demand response control for a heat pump heated residential building. Building Simulation, 11(3): 533–547. DOI: https://doi.org/10.1007/s12273-017-0425-5
Bessa, R, Moreira, C, Silva, B and Matos, M. 2014. Handling renewable energy variability and uncertainty in power systems operation. WIREs Energy and Environment, 3: 156–178. DOI: https://doi.org/10.1002/wene.76
Brown, MA and Zhou, S. 2013. Smart-grid policies: An international review. WIREs Energy and Environment, 2: 121–139. DOI: https://doi.org/10.1002/wene.53
Cao, S, Hasan, A and Siren, K. 2013. On-site energy matching indices for buildings with energy conversion, storage and hybrid grid connections. Energy and Buildings, 64: 423–438. DOI: https://doi.org/10.1016/j.enbuild.2013.05.030
Galus, MD, Vayá, MG, Krause, T and Andersson, G. 2013. The role of electric vehicles in smart grids. WIREs Energy and Environment, 2: 384–400. DOI: https://doi.org/10.1002/wene.56
Gao, L, Wu, H, Jin, H and Yang, M. 2008. System study of combined cooling, heating and power system for ecoindustrial parks. International Journal of Energy Research, 32: 1107–1118. DOI: https://doi.org/10.1002/er.1448
Kayo, G, Hasan, A and Siren, K. 2014. Energy sharing and matching in different combinations of buildings, CHP capacities and operation strategy. Energy and Buildings, 82: 685–695. DOI: https://doi.org/10.1016/j.enbuild.2014.07.077
Lund, P. 2012. Large-scale urban renewable electricity schemes – Integration and interfacing aspects. Energy Conversion and Management 63: 162–172. DOI: https://doi.org/10.1016/j.enconman.2012.01.037
Lund, PD. 2018. Capacity matching of storage to PV in a global frame with different loads profiles. Journal of Energy Storage, 18: 218–228. DOI: https://doi.org/10.1016/j.est.2018.04.030
Lund, PD, Lindgren, J, Mikkola, J and Salpakari, J. 2015. Review of energy system flexibility measures to enable high levels of variable renewable electricity. Renewable and Sustainable Energy Reviews, 45: 785–807. DOI: https://doi.org/10.1016/j.rser.2015.01.057
Lund, PD, Mikkola, J and Ypyä, J. 2015. Smart energy system design for large clean power schemes in urban areas. Journal of Cleaner Production, 103: 437–445. DOI: https://doi.org/10.1016/j.jclepro.2014.06.005
Mikkola, J and Lund, PD. 2016. Modeling flexibility and optimal use of existing power plants with large-scale variable renewable power schemes. Energy, 112: 364–375. DOI: https://doi.org/10.1016/j.energy.2016.06.082
Pilpola, S and Lund, PD. 2018. Effect of major policy disruptions in energy system transition: Case Finland. Energy Policy, 116: 323–336. DOI: https://doi.org/10.1016/j.enpol.2018.02.028
Sharif, A, Almansoori, A, Fowler, M, Elkamel, A and Alrafea, K. 2014. Design of an energy hub based on natural gas and renewable energy sources. International Journal of Energy Research, 38: 363–373. DOI: https://doi.org/10.1002/er.3050
Statistics Finland. 2017. Energy table service. Available at: http://pxweb2.stat.fi/sahkoiset_julkaisut/energia2015/html/engl0000.htm [Last accessed 21 August 2016].
Statistics Finland. 2018. Cold weather raised energy consumption in housing in 2016, 1 February 2018. Available at: https://www.stat.fi/til/asen/2016/asen_2016_2017-11-17_tie_001_en.html [Last accessed 25 June 2018].