Modelling West Whins
Modelling West Whins with PyLESA allows two outcomes for the project: the design and demonstration of applicable smart controls and identification of future system designs. In both cases the model must be representative and usable to provide a basis for these further analyses. First, the inputs implemented into PyLESA have been described with a special focus on the calibration to best model West Whins. Then, the outputs of the model have been calibrated to describe the current system of West Whins and identify the possible ways of improvement. A dedicated section deals with the validation of the model to see if it is consistent with data provided by external sources on West Whins. Finally, as the first users of PyLESA, a review of the modelling tool has been presented before summarizing the main outcomes of this section.
Configuring Inputs
This section explains how the Excel inputs file used in PyLESA has been created to model West Whins. This inputs file is available to download and have all the following points considered below.
​
​
Selected Control Option
PyLESA provides two options for controls: Fixed Order Controls (FOC) and Model Predictive Control (MPC). Simply put, the FOC only organises how demand is met based on pre-set rules with fixed levels of priority, whereas the MPC optimises the controls over a pre-set prediction horizon. More details are given on these two approaches the proposed smart controls.
The current controls in West Whins are only composed of setpoint temperature values on the Domestic Hot Water (DHW) storage and the DHW demand has a priority compared to the Space Heating (SH) demand. Besides this, there are no other controls implemented in West Whins. Consequently, when there is an energy demand from the dwellings, the energy centre provides the energy required, no matter the current sources available. Thus, the controls are not optimised, which means that a FOC is more suitable to model the current controls in West Whins and it was the one chosen in the model.
Demand Profiles
The demand profiles are important components of the model as it represents the energy needs from the dwellings. Besides the space heating and the domestic hot water demands provided by the energy centre, the demand also considers the non-heating electricity, which is required for the household lights and appliances.
Electrical Demand
For the electrical demand, several options have been investigated such as the use of the software HOMER which allows the generation of a profile based on different types of houses. (Stiel & Skyllas-Kazacos, 2012) However, we chose to use a different solution that did not require the use of a software - a generic profile from Strathclyde's Energy Systems Research Unit (ESRU) database (available in the resources section). (University of Strathclyde, ESRU, n.d.)
.png)
Figure 1 :Electrical demand profile for a generic house in 1989 taken from the ESRU database over the year
The electrical demand profile shown in Figure 1 has been used in the West Whins model. Although this considers buildings that are older than that of West Whins, this demand does not consider the electricity used for space heating. Therefore, by assuming the same occupancy, this electricity consumption is expected to reflect West Whins.
Space Heating Demand
The approach for the space heating demand differs from the electrical demand. A generic profile from the ESRU database has again been used, but this has also been calibrated with real data from Emoncms.
Figure 2 : Space heating demand of each dwelling of West Whins over a month
In Figure 2, the real space heating demand of each dwelling of West Whins is shown from the 14th of March to the 14th of April. Before the starting date, there are no values for the space heating. Due to the high variability of the space heating demand over the seasons, it is not possible to extend this real profile to the rest of the year. Therefore, it has only been used to calibrate the ESRU generic profile.
As the generic profile has peak values that are around seven times higher than the real demand, the first step of the calibration focuses on the range of values. This is as expected seeing as this generic profile considers houses built in 1989. These require considerably more space heating than eco-friendly dwellings due to their outdated efficiency.
The second step is to replace periods of minimal space heating with zero values. The real profiles have some periods of no heating, which is not the case for the generic profile. This is due to the hysteresis of heating the West Whins dwellings, which have a better thermal insulation that ensures longer lasting thermal comfort in comparison to the generic profiles.
.png)
Figure 3: Calibrated space heating demand profile over the year
Here, the calibration method explained above has been extended throughout the rest of the year and the calibrated space heating demand profile is shown in Figure 3.
Domestic Hot Water Demand
For the domestic hot water demand, a profile can be directly created using the data available within Emoncms. The Emoncms monitoring data has been recording the total volume of hot water since the 15th of December. It is assumed that there is no seasonal variation for the hot water demand., This hourly volume reading can be converted into energy consumption and extended for a whole year such that a hot water energy demand profile for the whole year is generated. The following formula has been used:
​
​
​
with ρ the water density, Cp the specific heat of water, and ΔT the temperature difference between input and output.
.png)
Figure 4 : Domestic hot water energy demand profile over a week
In Figure 4, the hot water energy demand profile over a week for the six West Whins dwellings is shown. This profile is interesting to analyse as we can see several peaks each day. The highest peaks of the day are in general during the morning or in the evening. However, the presence of several peaks can be explained by the different routines across each dwelling. For instance, some of the residents wake up and shower later, while others will take their shower in the evening.
Weather Resources
The weather resources, including the wind speed, the pressure, and the air temperature were sent by Graeme Flett, and show the local conditions of Findhorn from 2020. The source of this data was DarkSky which gives the current weather conditions of a defined location. (DarkSky, n.d.) Solar data was generated by the typical meteorogical year tool available on the european comission, photovoltaic geographical information system website. (EU, 2021)
​
Wind Turbine Generation
Findhorn has a local generation of electricity produced on site via three Vestas V29 wind turbines with a nominal power of 225 kW and an older Vestas V17 wind turbine with a nominal power of 75 kW. The characteristics of these turbines have been used to fill the PyLESA wind inputs. (Vestas, 1996), (Vestas, n.d.) However, only one type of wind turbine can be modelled in PyLESA at the same time. To overcome this constraint, the Vestas V17 is considered as a third of a Vestas V29 due to the nominal power ratio of the two turbines. 
This generation is designed for the whole Findhorn eco-village, leading to peak values up to 100 times higher than that of the demand profiles. As a result, the generation is scaled down to only consider the fraction used for West Whins. Firstly, it is assumed that each dwelling in Findhorn receives the same quantity of the wind turbine energy. In practice, we used the following sizing factor:
​
​
​
Using this sizing factor and the weather resources previously defined, the wind turbines generation for West Whins has been plotted on Figure 5.
.png)
.png)
Figure 5 : Wind turbines generation over the year after sizing factor
Heat Pump Model
Among the heat pump models provided by PyLESA, the generic regression air source to water heat pump model with a capacity of 14kW is chosen. Indeed, compared to the simple and Lorentz models, it allows a variation in the COP value, which is consistent with the reality. Figure 6 shows the heat pump COP variation of the West Whins energy centre via the Emoncms monitoring feeds.
Figure 6 : COP value for the West Whins heat pump over two weeks
Thermal Storage
The characteristics of the thermal storage have been filled using the ETP report. (Flett, et al., 2020) In the model of West Whins, only the 550L domestic hot water storage tank has been considered. The 100L buffer storage for space heating has been neglected. Further details will be given in the limitations of the model.
Grid Tariffs
The Economy 7 grid tariff is currently used in West Whins. This has two prices for imports and one for exports. For the imports, there is a low-cost period overnight at 115.04 £/MWh and a high-cost period during the rest of the day at 136.74 £/MWh. The exports tariff is fixed at 45.40 £/MWh. (Flett, et al., 2020) These prices have been directly integrated into the PyLESA inputs, as represented on the Figure 7.
Figure 7 : Economy 7 tariff used in West Whins over a day
Analysis of the Outputs
As the inputs have been defined within the previous section, it is now possible to run a simulation, using the West Whins model, for the whole year. In this section, key results are analysed. All values are extracted directly from PyLESA and can be replicated by running the West Whins model (available in the Downloads section).
Renewable Energy Usage
The total wind turbine energy generation throughout the year is over 11,841 kWh. This usage is shown in Figure 8. When there is wind, first the energy from the wind turbines is used to fill the electrical demand, then it supplies the heat pump electricity demand to meet the space heating and the domestic hot water demands, and finally the surplus is exported. Over the year, more than 6,117 kWh of electricity from the wind turbine generation is exported. This means that over half of the total wind turbine generation is being exported throughout the year.
.png)
Figure 8 : Wind turbines energy usage over the year
Grid Interactions
In Figure 9, the interactions with the grid have been plotted over the year. The annual imports are about 8,130 kWh while the exports represent 6,177 kWh for the whole year. In other words, the system is importing 57% of the time throughout the year. These imports occur during periods of no wind generation - when there is not enough energy from the wind turbine generation to supply the demand. This result underlines the necessity of implementing controls to ensure a better use of the wind turbine generation and to reduce imports.
.png)
Figure 9 : Grid interactions over the year
Cashflow
As the West Whins system is importing most of the time, there is a high cost of electricity. On the Figure 10, the cashflows have been plotted over the year. The benefits of the exports clearly do not balance the costs of imports. This financial aspect provides an additional justification for better controls.
.png)
Figure 10 : Cashflow over the year
Validation of the Model
This section focuses on the validation of the model to check if it is consistent with the current design of West Whins.
Space Heating and Domestic Hot Water Demands Profiles
For the West Whins PyLESA model, the total space heating and domestic hot water demands are respectively 12,537 kWh and 5,072 kWh for the whole year. In 2018, previous research shows the annual space heating and domestic hot water demands for the 6 flats are 11,577 kWh and 4,424 kWh respectively. (Flett, et al., 2020) This leads to a difference of a just over 9% for the space heating demand and under 15% for the domestic hot water demand between the model and the real design. Therefore, despite being a bit oversized, the demand profiles in the model show consistent values compared to the real consumptions in 2018.
Wind Turbines Generation Sizing Factor
The wind turbines generation directly impacts the interactions with the grid, as higher/lower generation reduces/increases imports and increases/reduces exports. Indeed, the assumption made on the sizing factor of the wind turbines generation will significantly impact the results. However, in 2018, for the whole Findhorn ecovillage, 659 MWh were imported, and 501 MWh were exported. (Flett, et al., 2020) As a result, the ratio of imports over the grid interactions was about 56.81% for 2018, still for the whole ecovillage. In our model of West Whins, this ratio is equal to 57.06%, which means that the part of imports of our model is close to the one of the ecovillage of Findhorn in 2018. Therefore, the assumption made on the sizing factor for the wind turbines generation seems reasonable.
Current Controls of West Whins
The current controls of West Whins are consistent with the ones defined in our model. Indeed, when there is a demand, it is directly being met, no matters if there is wind turbines generation. No forecasting is used to optimises the storage of energy to be used after. This is the case both in the fixed order control and in the current design controls of West Whins. However, in the real system, there is a priority between domestic hot water and the space heating demands that is not considered in our model.
COP of the Heat Pump
The data on the COP value, available with Emoncms was not enough to provide a complete validation for the whole year. However, the average COP value between the 20th of March 2021 and 12th of April 2021 was 2.11. (Pérez, 2021) This value is a bit under the average COP value of our model for this period, which is close to 3. Nevertheless, the value falls into the typical range for air source heat pump COP value, which is between 2 to 4. (Kelly, et al., 2010)