Electricity is a basic need in today`s society, being present in almost every aspect of life. The extensive infrastructure that transports electricity to the consumers, is expected to always work. At the same time there is a growing realism that the energy sector needs to modernise and ‘greenify’, in order to become more climate friendly. This is realised in the Netherlands by increasing the electricity produced by rooftop solar panels and wind parks.
Electricity produced by solar and wind depends on weather circumstances. Their production pattern is much more variable and uncertain than conventional fossil power plants. In order to maintain the balance on the electricity grid, an increased amount of balancing power is needed. For the Netherlands this is estimated to reach 15.2 TWh in the year 2050.
New sources of balancing power are also established by becoming more climate friendly. In the Netherlands the electrification of demand, results in gas fired boilers being replaced by heat pumps. Not just the established ground water based variant, but also the air based heat pump.
The heat pumps capability of providing balancing power is based on two aspects. First are the well-insulated houses in which they are installed. This reduces the cooling rate. Secondly is that residents generally do not experience any discomfort when the room temperature slightly deviates from their comfort temperature. And when they are absent, the room temperature may take any value.
Among the houses and residents where a heat pump is installed, differences can be found. Not everyone lives in a detached house, and not everyone is absent during the day. The effect of these differences on the demand response from a pool of heat pumps, is not known. This leads to the main research question:
“To what extent can a distinction in different residents and houses, increase the accuracy of demand response calculations for a Dutch heat pump pool?”
In order to answer the research question, a model based approach is chosen. The ISO 13790 modelling approach is selected, as it includes the following requirements:
1.Includes the comfort temperature of the resident
2.Includes the 4 MW participation requirement of the secondary reserve market
3.Allows the comfort temperature to change on an hourly basis
4.Allows for a distinction between different houses and residents
5.Includes the room temperature
6.Includes the air based heat pump`s dependency on the outside temperature
As a result of the 4 MW participation requirement, heat pumps need to be combined into a pool. An aggregator is tasked with controlling this pool, and offering the demand response to the market. It is the aggregator that must predict the demand response available from each heat pump. Distinguishing between the different houses and residents, should increase the accuracy of predictions.
The distinction between the different houses and residents programmed into the model, depends strongly on the information available about the region. The region chosen in this research is the Netherlands. For this region four different categories of houses have been identified, each with a different cooling rate.
•Detached house
•Semi-detached house
•Terraced house middle
•Terraced house corner
In case of the residents three different types have been identified. Each having a different comfort temperature and a different daily routine.
•The single
•The couple
•The family
Each of the three types of residents could live in any of the four categories of houses. All of the possible combinations have been included in the experiments. In the experiments the demand response available from the two types of heat pump has been calculated, in both the up- and downward direction. The demand response is here defined as the intentional deviation in the electric load, compared to the normal use of the heat pump.
The results of the experiments show that both types of heat pump are capable of keeping the house near to the comfort temperature. Between the two types of heat pump a significant difference was found. The air based variant is capable of providing more demand response, especially during the winter.
A significant difference was also found between the three different types of residents, and almost all of the four different categories of houses. With the exception of the semi-detached house and the terraced house middle, where no significant difference was found in the upward direction.
The conclusion is that by making a distinction between residents, houses and heat pumps the accuracy of calculations will increase.
The differences in the demand response available from each combination, effect the number of heat pumps needed for the 4 MW participation requirement. Depending on the time of day, the size of the pool would range between less than 10.000, and more than 180.000. It can be stated that it is not about finding the size of the pool that allows an aggregator to always participate. But it is about determining when an aggregator is capable of participating, based upon the pool he controls.
With the growing balancing power needed by 2050, heat pumps will be capable of providing part of the needed capacity. However, the balancing power offered fluctuates strongly depending on the time of day and season. Next to that, it will require large numbers of heat pumps in order to participate.