INOM
EXAMENSARBETE ENERGI OCH MILJÖ, AVANCERAD NIVÅ, 30 HP
STOCKHOLM SVERIGE 2020,
Techno-economic Potential of
Customer Flexibility, a Case Study
MARYAN BOURALEH
KTH
SKOLAN FÖR INDUSTRIELL TEKNIK OCH MANAGEMENT
Abstract
District heating plays a major role in the Swedish energy system. It is deemed a renewable energy source and is the main provider for multi-family dwellings with 90 %. Although the district heating fuel mix consists of majority renewables, a share of 5 % is provided from fossil fuels. To reduce fossil fuel usage and eradicate CO2-emissions from the district heating system new solutions are sought after. In this project, the potential for short- term thermal energy storage in buildings is investigated. This concept is referred to as customer flexibility.
Demand flexibility is created in the district heating system (DHS) by varying the indoor temperature in multi-family dwellings with maximum 1◦C, without jeopardizing the ther- mal comfort for the tenants. The flexible load makes it possible to store energy short-term in the building’ envelope. Consequently, heat load curves are evened in production. This leads to a reduction of the peak load in the DHS. Peaks are associated with high costs and negative environmental impact. Therefore, the potential benefits of customer flexibility are reduced peak production, fuel costs, and CO2-emissions, depending on the fuel mix in the DHS.
The project objective is to examine the techno-economic potential of customer flexibility in a specific DHS. The case study is made in a DHS owned by the company Vattenfall, located in the Stockholm area. To evaluate the potential benefits of implementing the concept, seven key performance indicators are chosen. They are peak power, peak fuel usage, produced volume, total fuel cost, fuel cost per produced MWh, climate footprint, and primary energy. Moreover, an in-house optimization model is used to simulate multiple scenarios of the DHS. Customer flexibility is modeled as virtual heat storage that can be charged up or down depending on the speed and size of the available storage at a specific outdoor temperature. The master thesis also aims to validate assumptions and parameters made in the input data to the optimization model.
Simulation results give a maximum peak power reduction of 10.9 % and annual fuel cost reduction between 0.9-3.6 % depending on the scenario. The results found are comparable to values found in similar studies. However, the environmental key performance indicators generated an increase in CO2-emissions and primary energy compared to the baseline sce- narios. The result would have looked different if fossil fuels were used in peak production instead of biofuels. The assumptions made in the input data to the optimization model was validated using results attained from a pilot in the specific DHS. Therefore results generated from the simulations are deemed accurate and confirm that customer flexibility leads to reduced peak production and DHS optimization.
Key Words: district heating, demand-side management, customer flexibility, thermal inertia, virtual storage, thermal energy storage, load shifting, space heating, multi-family dwellings, demand response
Sammanfattning
Fj¨arrv¨arme spelar en viktig roll i det svenska energisystemet. Den anses vara en f¨ornybar energik¨alla och ¨ar den huvudsakliga leverant¨oren f¨or flerfamiljshus med 90%. ¨Aven om br¨anslemixen i fj¨arrv¨arme sektorn best˚ar av f¨ornybara energik¨allor till en majoritet, utg¨or fossila br¨anslen fortfarande en 5% andel. F¨or att minska anv¨andningen av fossila br¨anslen och CO2-utsl¨app fr˚an fj¨arrv¨armesystemet efterfr˚agas nya l¨osningar. I detta projekt un- ders¨oks potentialen f¨or kortvarig v¨armelagring i byggnader. Konceptet kallas kundflexibi- litet i detta arbete.
Efterfr˚ageflexibilitet skapas i fj¨arrv¨armen¨atet genom att variera inomhustemperaturen i flerfamiljshus med maximum 1◦C, utan att p˚averka inomhusklimatet f¨or de innebo- ende. Den flexibla lasten g¨or det m¨ojligt att kortsiktigt lagra energi i byggnadens kli- matskal. F¨oljaktligen j¨amnas v¨armelastkurvorna i produktionen. Detta leder till en minsk- ning av topplasten i fj¨arrv¨armen¨atet. Effekttoppar ¨ar f¨orknippade med h¨oga kostnader och en negativ milj¨op˚averkan. D¨arf¨or ¨ar potentiella f¨ordelar med kundflexibilitet redu- cerad toppproduktion, br¨ansle-kostnader och CO2-utsl¨app, beroende p˚a br¨anslemixen i fj¨arrv¨armen¨atet.
Projektets m˚al ¨ar att unders¨oka den tekno-ekonomiska potentialen f¨or kundflexibilitet i ett specifikt fj¨arrv¨armen¨at. Fallstudien g¨ors i ett Vattenfall-¨agt fj¨arrv¨armen¨at lokali- serat i Stockholmsomr˚adet. F¨or att utv¨ardera och analysera de potentiella f¨ordelarna med att implementera konceptet v¨aljs sju resultatindikatorer. De ¨ar: toppeffekt, topp br¨anslef¨orbrukning, producerad volym, total br¨anslekostnad, br¨anslekostnad per produ- cerad MWh, klimatavtryck samt prim¨ar energi anv¨andning. En intern optimeringsmodell anv¨ands f¨or att simulera olika scenarier av fj¨arrv¨armen¨atet. Kundflexibiliteten modelleras som ett virtuellt batteri som kan laddas p˚a eller ur beroende p˚a batteriets hastighet och storlek vid en specifik utomhustemperatur. Examensarbetet syftar ocks˚a till att validera antaganden och parametrar som anv¨andes i inmatningsdata till optimeringsmodellen.
Simuleringsresultat ger en maximal toppeffektreducering med 10.9 % och en ˚arliga br¨ansle- kostnadsbesparing mellan 0.9-3.6% beroende p˚a scenario. Resultaten ¨ar j¨amf¨orbara med v¨arden funna i liknande studier. Nyckeltalen g¨allande klimatavtrycket och prim¨ar energi- anv¨andningen i de olika scenarierna som simulerades genererade en ¨okning j¨amf¨ort med referensscenarierna, d¨ar ingen kundflexibilitet antogs. Resultaten skulle sett annorlunda ut om fossila br¨anslen anv¨andes i topplastproduktion ist¨allet f¨or biobr¨anslen. Antaganden om den tillg¨angliga flexibiliteten i fj¨arrv¨armen¨atet validerades genom att anv¨anda pilot- resultat gjorda i det specifika fj¨arrv¨armen¨atet. D¨arf¨or anses resultat fr˚an simuleringarna vara korrekta och bekr¨aftar att kundflexibilitet leder till en minskad toppproduktion och n¨atoptimering.
Nyckelord: fj¨arrv¨arme, efterfr˚agesidahantering, kundflexibilitet, termisk tr¨oghet, virtuell lagring, v¨armeenergilagring, lastf¨orskjutning, byggnadsuppv¨armning, flerfamiljshus, de- mand response
Acknowledgements
After handing in this master thesis project I have completed 5 years of the Mechanical Engineering program and two years of master studies in the Sustainable Energy Technol- ogy program at KTH Royal Institute of Technology. For achieving this feat I would like to profess immense gratitude to my family for their endless encouragement. Thank you for keeping me driven and enriching my life. I would also like to mention a few people that have helped me in my master thesis.
To start, I would like to express my gratitude to Bj¨orn Franck and Cecilia Ib´anez- S¨orenson for introducing me to Vattenfall the summer of 2019 and allowing me to grow professionally.
Secondly, I send many thanks to Christina Hyllander for getting me in contact with Vattenfall R&D to do my master thesis and providing me with an amazing opportunity.
I did my thesis in the R&D team Energy System Optimization and would like to thank everyone for being so welcoming, it was a pleasure working with you. I am especially indebted to Shahriar Badiei and Sofia Petersson Svanfeldt for helping me validate assumptions and being open for further discussions.
Moreover, I would like to send sincere appreciation to Wouter Verbeeck for helping me understand and modulate input data to the optimization model.
Mostly I would like to thank my supervisor at Vattenfall, Fanny Lindberg, and my supervisor and examiner at KTH Jaime Arias Hurtado. Fanny, thank you for your constant support throughout the whole project and for helping me develop professional skills by observing your work ethic. Jaime, thank you for your curious nature and for helping provide insight from a building’s perspective.
Table of Contents
1 Introduction 1
1.1 Thermal Energy Storage . . . 3
1.2 Vattenfall . . . 4
1.3 Aim & Objectives . . . 4
1.4 Methodology . . . 5
1.4.1 Approach . . . 5
1.5 Limitations . . . 6
2 Customer Flexibility 7 2.1 Applications . . . 7
2.2 The Available Customer Flexibility . . . 8
2.2.1 Thermal Comfort . . . 9
2.2.2 The Building . . . 9
2.2.3 The District Heating System . . . 13
2.3 Other Studies . . . 13
3 Case Study 15 4 Modeling 18 4.1 C3PO . . . 18
4.2 Virtual Storage . . . 18
4.2.1 Net Zero Energy . . . 20
4.3 Key Assumptions . . . 21
4.4 Two Optimization Strategies . . . 22
4.4.1 Normal-operation Optimization . . . 22
4.4.2 Peak-operation Optimization . . . 23
4.5 Scenarios . . . 23
4.6 Calculating The KPI Values . . . 24
4.6.1 Peak Power . . . 25
4.6.2 Peak Fuel Usage . . . 25
4.6.3 Produced Volume . . . 25
4.6.4 Total Fuel Cost . . . 25
4.6.5 Fuel Cost per MWh . . . 25
4.6.6 Climate Footprint . . . 25
4.6.7 Primary Energy . . . 26
5 Results 27 5.1 Normal-operation Optimization . . . 27
5.1.1 Baseline . . . 27
5.1.2 Normal Flexibility . . . 28
5.1.3 10 % Demand Reduction . . . 29
5.1.4 30 % Demand Reduction . . . 30
5.1.5 50 % Demand Reduction . . . 31
5.2 Peak-operation Optimization . . . 32
5.2.1 EOT5 Baseline . . . 33
5.2.2 EOT5 Normal Flexibility . . . 33
5.3 KPI Fulfillment . . . 34
6 Discussion 36 6.1 Results . . . 36
6.2 Key Assumptions . . . 37
6.3 C3PO . . . 39
7 Conclusions 40
8 Future Work 41
Appendix A Interpolation Table I
Appendix B KPI Values for Normal-operation Scenarios III
Appendix C KPI Values for Peak-operation Scenarios IV
Appendix D Climate Footprint & Primary Energy Calculations V
List of Figures
1.1 Supplied energy to the Swedish district heating system between 1980-2018 2
1.2 Fuel mix in the Swedish district heating system, 2018 [4] . . . 3
1.3 Examples of thermal energy storage [10] . . . 4
1.4 Master thesis approach . . . 6
2.1 Flow chart over heat journey in buildings . . . 7
2.2 Customer flexibility . . . 8
2.3 Deciding factors of customer flexibility . . . 9
2.4 Heat load for a typical weekday . . . 12
2.5 The variation of the daily heat load between seasons . . . 12
3.1 Outdoor temperatures, 2018 [26] . . . 15
3.2 DHS Fuel Mix . . . 16
4.1 An illustration of the virtual storage . . . 19
4.2 Energy signature for 50 largest MFD in the DHS . . . 19
4.3 Mathematical model of the virtual storage when NZE is rejected . . . 21
4.4 Normal-operation optimization . . . 23
4.5 Peak-operation optimization . . . 23
5.1 Heat production curve for Baseline scenario . . . 28
5.2 Heat production curve for Normal Flexibility scenario . . . 28
5.3 Heat production curve for 10 % Demand Reduction scenario . . . 30
5.4 Heat production curve for 30 % Demand Reduction scenario . . . 31
5.5 Heat production curve for 50 % Demand Reduction scenario . . . 32
5.6 Heat production curve for EOT5 Baseline scenario . . . 33
5.7 Heat production curve for EOT5 Normal Flexibility scenario . . . 33
List of Tables
2.1 Thermal inertia of some materials . . . 10
4.1 The values of the assumption assumed . . . 22
4.2 Definition of scenarios for normal-operation optimization . . . 24
4.3 Definition of scenarios for peak-operation optimization . . . 24
5.1 KPI values for Normal Flexibility scenario . . . 29
5.2 KPI values for 10 % Demand Reduction scenario . . . 30
5.3 KPI values for 30 % Demand Reduction scenario . . . 31
5.4 KPI values for 50 % Demand Reduction scenario . . . 32
5.5 KPI values for EOT5 Normal Flexibility scenario . . . 34
5.6 KPI fulfillment for all scenarios . . . 34 A.1 Interpolation table . . . I B.1 KPI values for all normal-operation scenarios . . . III C.1 KPI values for all pormal-operation scenarios . . . IV D.1 Key values for climate footprint and primary energy calculations . . . V
Nomenclature
Acronyms
CHP Combined Heat and Power
DH District Heating
DHS District Heating System
DHW Domestic Hot Water
HOB Heat Only Boilers
KPI Key Performance Indicator
MFD Multi-Family Dwellings
NZE Net Zero Energy
TES Thermal Energy Storage
Greek Letters
λ Heat transfer coefficient of a specific material [W/m2K]
ρ Density [kg/m3]
τ Time constant, defines the thermal inertia of a building [h]
Physical constants
c Specific heat [kJ/kg-K]
I Thermal inertia of a material [W s0.5/(M/K)]
m Mass [kg]
t Discharging time [h]
Other Symbols
∆Ti The indoor temperature variation [◦C]
∆To The outdoor temperature variation [◦C]
Cth Thermal mass of a specific building [W/K]
Qgain The heat gains in a building [kWh]
Qheat The heating required in a building [kWh]
Qloss The heating losses in a building [kWh]
Chapter 1 Introduction
The apparent climate change and natural disasters that are occurring more often world- wide is a reminder that global warming is evident and must be counteracted [1]. To combat the ever-changing environment and provide a stable future, continuous devel- opment must be sustainable. To succeed in that aim, key factors are the reduction of greenhouse gases, a greater share of green power in energy production, and overall energy efficiency [2]. These are all goals set by the UN and are targeted for 2030.
The EU has renewed its energy policy framework since the Paris Agreement in 2015 to a new rulebook by the name Clean energy for all Europeans package. This entails; improved energy performance in buildings, who is the biggest energy-consuming sector in Europe, a higher renewable energy share, and energy efficiency which aims to reach EU:s long strategy goals to be carbon neutral in 2050 [3]. Part of the new rulebook is a new energy efficiency directive that was updated in 2018 after the last one in 2012. The new directive says that all EU members must reach energy savings of 0.8 % each year from their energy consumption between 2021 and 2030.
In Sweden, the total energy consumption was 378 TWh in 2017. The largest consumption came from the residential and service sector, buildings, with a share of scarcely 40 %. The majority of the energy used in the residential sector is for heating. Moreover, the largest share of the heat supply with 70 % goes to space heating and domestic hot water (DHW) usage of residential and non-residential buildings. 60 % of these are supplied to using district heating (DH). In multi-family dwellings (MFD) DH is the main energy provider with 90 % [4]. DH is, therefore, a central energy carrier in Sweden’s overall energy system.
Figure 1.1 below shows the supplied energy to the district heating systems (DHS) between 1980 and 2018.
Figure 1.1: Supplied energy to the Swedish district heating system between 1980-2018 [5]
Given the importance of DH in the Swedish energy system, the supplied energy to the DHS is also of significance. The majority of the fuel used is renewable or recycled energy.
It should be noted, however, that there are some traces of fossil fuels left in the fuel mix.
The share of fossil fuels in Sweden was 5 % in 2018. Still, the CO2-emission rates have reduced drastically since 1980 but have been rather constant in the last five years even with an increase in heating consumption because of a growing population [6]. This is an immense accomplishment. The EU target, however, is to reduce the CO2-emission, all greenhouse gases, with 40 % by 2030 and have net-zero green house gas emissions by 2045 [7]. See figure 1.2 for Sweden’s fuel mix in the DHS in 2018.
Figure 1.2: Fuel mix in the Swedish district heating system, 2018 [4]
The fuel mix mostly consists of biomass and waste. Other sources of heat are heat pumps and industrial waste heat. However, roughly 5 % of the supplied energy to the DHS comes from fossil fuels. To decrease the dependency on fossil fuels, considering conversion to more sustainable fuels is an immediate option. To accompany the complete conversion to new fuels other complementary solutions are considered.
1.1 Thermal Energy Storage
To reduce fossil fuel usage in DH production and simultaneously increase the energy efficiency in buildings, thermal energy storage (TES) in buildings has been the topic of discussion. The heat loads in DHS vary because of the weather and customer use, which makes it difficult to match the demand from customers and optimize the heat system [8]. TES can be used to create flexibility in the DHS and there exist various ways to do so. Firstly, it should be explained that TES can be divided into sensible and latent heat storage. Latent heat is for storage in phase change materials. Sensible heat storage increases temperature without phase changing a material [9]. Under this category follows many subcategories, see figure 1.3 below.
Figure 1.3: Examples of thermal energy storage [10]
Centralized TES is a hot water storage tank that is connected to the entire DHS and acts as a buffer when in-need. The tank stores excess heat from the whole system and can be charged or discharged. TES in the distribution DHS is a second way to create flexibility in the DHS. It is possible by adjusting the supply temperature. However, distribution losses increase with increased water temperature. Borehole storage is an effective way to store heat between seasons. It is possible by storing heat deep down in the summer for re-use in the winter when the heat is more in-demand and therefore more expensive [10]. The last example of TES presented utilizes the thermal inertia of buildings to store heat and is a short-term TES. The concept is referred to as customer flexibility and is of interest in this master thesis report.
1.2 Vattenfall
This master thesis is a collaboration with Vattenfall. Vattenfall is a 100 % Swedish state-owned energy company with 20 000 employees and markets in Sweden, Denmark, Germany, Netherlands, the UK, Finland, and France. Vattenfall is one of the largest energy corporations in Europe [11]. The company is active in the entire energy value chain which includes production, distribution, trading, retail, and services. The company’s purpose is to Power Climate Smarter Living and therefore be leading in sustainable production and consumption. Moreover, Vattenfall’s vision is to be fossil-free within one generation [12]. Therefore, the company focuses on finding solutions to reduce its fossil fuel need. In the Nordic, the goal is to be completely climate-neutral in energy production by 2025 [13]. Additionally, the Government of Sweden wants to be climate neutral and therefore fossil-free by 2045 [14]. These are both driving forces to make Vattenfall more energy-efficient and climate neutral.
1.3 Aim & Objectives
This project has the objective to examine the techno-economic potential of customer flex- ibility in a specific DHS. To evaluate the potential benefits of implementing short-term building TES in the DHS, an in-house optimization model is used. Multiple scenarios are created with different parameter values to simulate results using different sets of assump- tions about the available flexibility, the thermal characteristics of the buildings, and the DHS. The driving forces for this master thesis are overall energy system optimization as
well as fossil-free district heating. The aim is therefore to reduce fossil-free usage, mostly used in peak production.
The research questions chosen for the master thesis project are the following:
1. How will the utilization of customer flexibility affect the following key performance indicators (KPI)?
• Peak power [MW]
• Peak fuel usage [MWh]
• Produced volume [MWh]
• Total fuel cost [SEK]
• Fuel cost per MWh [SEK/MWh]
• Climate footprint [tCO2-e]
• Primary energy [MWh]
2. How can assumptions and parameters used in the optimization model be validated?
1.4 Methodology
The research methodology used in this project was both a qualitative and a quantitative one. A literature study was conducted to give a theoretical background regarding the subject and exploring the available information. Focus areas were district heating, thermal inertia in buildings, and TES in buildings. The primary sources were KTH Library (Primo), Google Scholar and internal documents. Secondly, a quantitative method was used by using an optimization model and measuring flexibility in terms of predetermined KPIs. The qualitative modeling included an analysis of the input data, in terms of temperature data, heat demand, and flexibility availability. The analysis also continues after scenarios have been simulated to investigate the effects of the parameter changes on the KPI results.
1.4.1 Approach
In figure 1.4 below, the approach used in the project is presented. It consists of a six-step process. Firstly, research was made to introduce the student into the subject area and help structure the project. Moreover, the scope and key questions were defined. Secondly, a literature study was conducted to give the author useful information about the subject of demand flexibility. The literature study was an essential step in understanding and fully comprehending the project and its underlying theory. It included a theoretical background about district heating, thermal characteristics about buildings, and finding similar studies. Thirdly, the available optimization model was reviewed. Based on the literature study, new parameters could be implemented and the model could be redefined resulting in new simulations and new results to analyze. The process between step 3-4 is iterative and was continuous throughout the project to investigate the effect of changing parameters on the chosen KPIs. In step 5 the best scenarios are chosen and declared the final result. The results and methodology are later discussed. Lastly, conclusions from the master thesis are drawn and possible deliverables are presented as a last step in the project.
Figure 1.4: Master thesis approach
1.5 Limitations
The project focuses mainly on one specific DHS and only the 50 largest multi-family dwellings in that DHS. The aim is therefore to optimize the DHS after the given seven KPIs and optimize the existing model. Moreover, the master thesis is investigating only the possibility to reduce the space heating demand and therefore does not consider DHW usage or district cooling. Furthermore, DHS using only heat plants are considered. There- fore, no electricity production is considered. Moreover, it is a limitation that the DHS chosen does not have fossil fuels in its energy mix.
This project report is also limited in privacy because the project is done for Vattenfall, some information is therefore withheld for company privacy. Moreover, the DHS investi- gated is restricted to company-owned ones.
Chapter 2
Customer Flexibility
This project has an objective to investigate customer flexibility in a DHS. This chapter gives a theoretical background and provides information from similar studies.
In figure 2.1 below, a flow chart for the heat in the building is shown, from production to customer. It is within these interactions that flexibility can be found and optimize the entire DHS. The heat is deemed to be in the DHS from production until it reaches the substation, then it is deemed to belongs to the building.
Figure 2.1: Flow chart over heat journey in buildings
The production looks different in many DHSs because of the fuel mix and their associ- ated boilers. The distribution system transports heat from the plant to the substations connected to the buildings, through a piping system causing heat losses [15]. The heat from the radiators heat the buildings and with time is stored into the building envelope [16].
Customer flexibility in this report is referred to as thermal energy storage in buildings using the thermal inertia of the DHS’s associated buildings. Given that buildings have an envelope with a certain mass, it takes time for heat to exhaust a building. This inertia can be taken advantage of and create flexibility in the DHS. The term is called customer flexibility because the flexibility is created by varying the indoor temperature of the MFDs. Customer flexibility works as temporary thermal storage which creates the possibility for loads to be shifted over a short period. This is also called load shifting [17].
2.1 Applications
Making use of the thermal inertia in buildings, a building can be viewed as virtual storage.
In the same way, a battery can be charged up or down, a building can be charged up or down with heat by varying the indoor temperature. Before peak, when the load is lower the virtual storage is charged up by increasing the indoor temperature in the building and thus storing heat in the building. This heat can later be used during peak hours,
thus the storage tank is discharged and loads are shifted over time, see figure 2.2 for an illustration of the concept.
Figure 2.2: Customer flexibility
There are multiple benefits with customer flexibility [10], some examples are:
• Even production curves
• Reduction of peak loads
• Reduction of boiler capacity
• Reduction of CO2-emissions
• Reduction of fuel costs
Creating flexibility in the DHS can balance out the heat demand from buildings and thereby lower the load when it is high and shift the load to time periods when the load is lower, this in turn evens out the heat production. This leads to a reduction of the peak loads in the DHS. Peaks are associated with high costs and environmental impact. For peak production usually, Heat-Only-Boilers (HOB) are used and they are for the most part fueled by fossils [18]. Reducing the peak will, therefore, lead to a possible reduction of these boilers and a reduction of CO2-emissions. With fewer boilers needed the operation and maintenance costs can also be reduced.
2.2 The Available Customer Flexibility
The potential of customer flexibility can be said to be limited by three major factors as seen in 2.3. The possibility to vary the indoor temperature in MFD is dependant on the thermal comfort for the tenants. Secondly, the building is an important factor. Its thermal characteristics decide energy storage availability. Moreover, the location of the building is important since it decides the weather and the heat losses of the building. Lastly, the DHS is an essential contributor to quantifying the potential of customer flexibility as well.
Figure 2.3: Deciding factors of customer flexibility
All three factors are discussed further below.
2.2.1 Thermal Comfort
An important factor to consider when utilizing the building envelope TES is the thermal comfort in the MFDs. To not negatively affect the thermal comfort in the building the maximum allowed predicted percentage dissatisfied is set to 10 %. That creates a limitation and results in a maximum temperature variation of maximum 1◦C from the set indoor temperature. That translates to a range of +/- 0.5◦C [19].
2.2.2 The Building
Multiple factors decide the TES availability of a building. Firstly, the thermal character- istics of that building are of importance. Moreover, the weather and climate in the region the MFDs are built in also affect the amount of heat that can be stored. Key concepts are presented below.
Thermal Inertia
Thermal inertia is an integral concept that must be explained to fully comprehend the project. Taking advantage of the thermal inertia of buildings will help create the customer flexibility that is aimed for. In essence, thermal inertia describes how a building responds to weather changes. The term is a material characteristic and is calculated using the thermal conductivity of the building material, its density, and heat capacity [20]. The formula for the term is described in equation 2.1 below.
I =p
λ · ρ · c (2.1)
Thus, the thermal inertia defines a material’s ability to absorb heat. Below in table 2.1, the thermal inertia of some materials is given.
Table 2.1: Thermal inertia of some materials Material Thermal Inertia [W s0.5/(M/K)] [20]
Mineral Wool 40
Concrete 1 800
Brick 900
Gypsum 400
Wood 310
In general, the lower the thermal inertia is, the easier it is for a building to absorb heat from the outside and thus affect the indoor temperature. For heavier buildings, it takes a longer time for the building to be affected by the outdoor temperature. Heavy buildings, external walls made of concrete, such as in MFDs generally have a time constant of 200 hours [16].
The time it takes for the building to reach the outdoor temperature, using no additional heating, can be a means to measure the thermal capacity of a building. The time constant τ described the time it takes for a building before it reaches 63 % of its end temperature, T∞. The time constant is used as a measure that shows how fast a building reacts to weather changes. The higher the time log, the more heat is required to change the buildings’ indoor temperature. The time constant is defined as the quota between the thermal mass of the building and its specific power demand[20].
τ = Cth
Qloss = P mjcj
Qloss (2.2)
Thermal mass is a material characteristic of the fabric used in the building envelope.
Given a light or heavy building material such as wood or concrete results in a low or high thermal mass. The heavier material that is used the longer it takes for the material to vary according to the outdoor temperature. A light timber-framed building varies according to the outdoor temperature and has a very low thermal storage capacity in comparison to the heavy building that barely changes its temperature over time [20]. The thermal mass is defined in equation 2.3 as the following:
Cth =X
mjcj (2.3)
Where m is the mass, c the specific heat of the material, and j the number of layers the building has. Summarizing the multiplication of the mass and specific heat of the material layers results in a total thermal mass of the building. The thickness of the wall equals higher mass and therefore higher thermal mass.
The specific power demand depends on the transmission, ventilation and infiltration losses as mentioned earlier. Moreover, materials with higher specific heat capacity generate a
greater time constant. Using the time constant which can be calculated for a specific house, the indoor temperature can be calculated as shown in equation 2.4 below [20].
∆Ti = ∆To· (1 − e−t/τ) (2.4)
Where ∆Ti is the temperature difference in indoor temperature and ∆To is the tem- perature difference between the outdoor temperature and a chosen balance temperature [20].
Heat Balance in Buildings
As mentioned previously in chapter 1, buildings are the most energy-consuming sector in Sweden. Moreover, this report focuses on using buildings as a TES. It is therefore essential to understand the energy balance in buildings. The heat that is consumed is the difference between the losses and the internal heat gains in a particular building, see equation 2.5. Losses consists of transmission, ventilation, infiltration, see equation 2.6.
The transmission losses comes from the building envelope, which are windows, doors, roof, walls, and floor. The gains come from heat emission from people, electrical appliances, and solar energy, which is of course weather and orientation-dependent, see equation 2.7.
The additional gains must be subtracted from the heat loss value, resulting in the amount of heat that must be supplemented to the system. Below are the governing equations:
Qheat= Qloss− Qgain (2.5)
Qloss = Qtrans+ Qvent+ Qinf (2.6)
Qgain = Qperson+ Qsolar + Qel.appl. (2.7)
The heat in the building is used for space heating and DHW. Space heating is weather dependent whereas the DHW usage is connected to customer behavior, it is usually set to be constant throughout the year but is a little higher in winter and little lower than average in the summer [16].
Heat Demand
The heat demand is the heat needed by the buildings to maintain a set indoor temperature at a specific hour. Depending on the weather and the customer activity the demand varies.
The heat load is dependant on the heat demand and describes the heat that is to be produced in total, considering distribution losses. It is therefore important to not confuse the two terms. The heat demand comes from the consumers and does not include the distribution losses. The heat load describes the production needed to cover the demand and distribution losses.
The heat demand is time-dependent and usually has two peaks during the day, as seen in figure 2.4. There are different loads; peak load and normal load. A peak is when the load is higher than the average load produced, i.e. normal operation. The normal operation load is the power demand that is in the DHS in most cases throughout the year. Peak load is the maximum power demand during a time-cycle and is usually much higher than the normal load. The peaks are usually in the morning before work and one later in
the evening. To illustrate this fact is figure 2.4 showing the heat demand for a typical weekday [18].
Figure 2.4: Heat load for a typical weekday
In figure 2.4, it is visible that the daily heat load has two larger peaks during the day as predicted.
The majority share of the heating demand comes from space heating, the demand is very weather dependent and is not consistent over the year. Below is figure 2.5 showing the daily heat load curve for a day in February, July, and October.
Figure 2.5: The variation of the daily heat load between seasons
Figure 2.5 shows that there are high peaks appear during the winter in February when the heating demand is high and the lowest loads are in the summer when there is almost no demand for heating at all, which is the case in July. In October, the load is closer to normal operation.
2.2.3 The District Heating System
The DHS is a crucial factor in quantifying the space heating flexibility, precisely because it depends on how the DHS is built, what the fuel mix consists of, and the type of buildings that are associated with it. The amount of power, size of the DHS, is dependant on the energy consumption of the buildings associated with the DHS. The available power decides the speed of the virtual storage that customer flexibility is modeled as. Moreover, the DHS is limited to its boiler capacity. It is also dimensioned after the design load.
Design Load
DHS uses a safety margin for its boiler capacity to assure heat supply to the buildings, this is referred to as the design load. Although the annual peak load could be a certain value, the boiler capacity must be dimensioned after the worst-case scenario. The reasoning behind that is that the heating consumption is weather dependent and can therefore never be entirely predicted. Therefore, the capacity in the power plants is dimensioned after the five coldest consecutive days in that specific region in the last 30 years . This is referred to as EOT5 and used to dimension the maximum load, boiler capacity, in the DHS [18]. Moreover, sometimes the largest boiler is dimensioned to be off simultaneously.
This means that even though the full capacity is not needed at all times the plants are over-dimensioned and are limited to scale down their boiler capacity because of the dimensioning load.
This project aims to investigate the potential for customer flexibility to reduce fossil fuel needs. The fossil fuels in DHS are mostly used in reverse boilers used to supply the peak loads in DHSs during the coldest days of the year. These boilers are used in combination with the base-load boiler capacity to cater to the additional heat load. Fossil fuel boilers are used because they are flexible and are handy with rapidly varying loads. The reserve boilers are often HOB and are added to the baseload boiler capacity when the heat demand is high, such as in cold winter periods or daily peak loads. Moreover, HOBs are used to start and stop production [21]. The easiest fuel to use when the boilers are being quickly started or stopped are fossils. Moreover, they are cheaper because they are not solids.
Mostly, it is oil and gas that are used [18]. For fossil-free production, the conversion to other fuels such as biofuels is an immediate option. Biofuels are however more expensive but have similar characteristics as fossils.
2.3 Other Studies
In this section the potential found about customer flexibility in literature is presented.
After all, terminology that helps define the benefits with customer flexibility is explained, the potential of it in reports and off-the-shelf optimization systems can be compared.
A key report in this study is The Value of Flexible Heat Demand written by the Swedish research and knowledge institute EnergiForsk. They focus on making the energy sector
smarter [22] . In this report, three possibilities to create flexibility in the DHS are com- pared and simulated. One of those is TES in buildings. An economic optimization was made using the flexibility of either 20 % or 44 % share of the buildings in an imaginary DHS with a total yearly consumption of 500 GWh. The boiler capacity consists of 6 HOBs that are used for peak production and have a capacity of 95 MW, whereas 80 MW of that is fueled by fossils. It is important to note that this share of fossils for HOBs is rather unrealistic. Using a simulation tool, economic optimization was made to investigate the potential of reducing operational costs with the help of customer flexibility. Three energy mixes were considered, they are majority combined heat and power (CHP), majority heat pumps and majority excess heat such as waste or industrial heat. The savings in opera- tional costs are 4.0 % if 20 % of the buildings are used as customer flexibility and 7.1 % if 44 % of the buildings are used. The allowed temperature variation in the building is also +/- 0.5◦C [10].
Another report that has been investigated is Smart Energy DHSs: Utilization of Space Heating Flexibility, written by Johan Kensby Ph.D. at Chalmers University and co-author of the EnergiForsk report. In this report, a simulation of the entire city of Gothenburg DHS was made using 20 % customer flexibility. Here it is calculated that 5.5-11 % of cost savings can be made from a reduction of fuel costs and revenues from sold electricity. The daily heat load variations also reduce with 50 % [17].
Lastly, in a ScienceDirect article written by Chalmers University in Gothenburg TES using thermal inertia in buildings is calculated to have a saving of 1 % on the annual operational costs It is worth noting that the results also include revenues from sold electricity because heat pumps are considered in the energy mix [23].
Also, because the reports only have resulted from model simulations it is interesting to see the calculated potential from available off the shelf solutions that are based on real values from customers.
NODA Smart-Heat Building uses indoor sensors and a smart control system to optimize the DHS and claims energy cost savings between 10-12 %. It is demand-side management control and reduces the supply temperature to the building based on the indoor temper- ature [24]. Ecopilot is another system that controls a building’s heating, cooling, and ventilation consumption in real-time. Promised savings are 22 % for heat. The system considers the building’s characteristics, internal loads, and sun exposure and energy prices [25].
Chapter 3 Case Study
The DHS chosen for a case study to measure customer flexibility is one located in the Stockholm area. The temperature varies quite a bit with low temperatures in the winter translating in a great need for heat and high temperatures in the summer where less heat is in-demand. See figure 3.1 below.
Figure 3.1: Outdoor temperatures, 2018 [26]
The DHS has about 400 associated buildings. The building types are multifamily houses, single-family houses, and facilities. The buildings that are going to be used as customer flexibility are the 50 largest energy-consuming MFDs, they represent 55 % of the total energy consumption in the DHS.
Multi-family dwellings are used as customer flexibility because are well suited for short- term building TES. They are usually built with heavier materials that have large thermal mass resulting in long time constants. Moreover, other benefits are that multi-family dwellings have similar heating systems. Lastly, they are the biggest consumer of DH in the residential sector [17].
The fuel mix for the DHS is show below in figure 3.2:
Figure 3.2: DHS Fuel Mix
The generated heat comes from majority wood-mix such as wood chips and pellets with 75 %. Renewable biofuels such as biogas, bio-oil, and biodiesel are also used. Bio-oil is used as a fuel for the peak boilers and has a share of 10 %. This share is aimed to be reduced.
The DHS utilized a total of 9 boilers:
• 1 wood chips boiler
• 1 biogas boiler
• 2 pellets boilers
• 2 bio-oil boilers, and
• 3 biodiesel boilers
Each boiler also has a boiler efficiency in production. The efficiency creates production losses. In the specific DHS, there are two boiler efficiencies considered. Maximum and minimum. The minimum boiler efficiency is lower for majority of the boilers compared to the maximum boiler efficiency. This means that using the maximum capacity of the boilers generates less production losses.
The total capacity of the boilers is dimensioned for EOT5 −16◦C. As seen, there are no fossil fuels in the fuel mix and in particular the peak production as reverse boilers to cover peaks in the specific DHS. However, a driving force for the project was fossil-free district heating and the aim was therefore to investigate the effect of customer flexibility on the peak load and in particular the fossil fuel usage. Therefore it could be concluded that the grid chosen contradicts the aim. The particular grid was chosen because there was
available data in terms of building consumption and it is also a test-bed grid. Since pilot studies are often made in the DHS, results from those pilot studies can be used to validate the assumptions made in the modeling of the virtual storage. Thereby making the results from the master thesis more reliable. Regardless, the objective is also to investigate the effect of customer flexibility on several KPIs that could all be calculated and analyzed.
Chapter 4 Modeling
In this chapter, the optimization model used to simulate customer flexibility is introduced.
Firstly, some background information about the model is provided. Secondly, the model- ing of customer flexibility in the in-house model is explained. Thirdly, the key assumptions made in the modeling and the scenarios chosen to simulate are presented. Furthermore, optimization strategies are explained. Lastly, the significance and the calculation of the KPI values are explained.
4.1 C3PO
Combined Power Production Portfolio Optimizer, referred to as C3PO, is an in-house op- timization model developed by Vattenfall. The Python-model uses mixed-integer-linear programming to optimize production by considering the power plants in the DHS, market fuel prices, DHS connections, and storage possibilities. Here, customer flexibility is sim- ulated as virtual heat storage. The model does an economic optimization and is specific to the chosen DHS.
The model uses historic data such as annual temperature and hourly heat demand data of building’ in the specific DHS. The data was provided for the year 2018. Moreover, distribution losses of 10 % in the DHS are assumed [15], which helps to calculate the amount of heat that needs to be produced to cover the building’s consumption. C3PO also considers the capacity, efficiency, and fuel type of the boilers. Files with information regarding each boiler, its fuel prices as well as information regarding the virtual storage are put into an interface that simulates the specific scenario. The information are such regarding the size and speed of the virtual storage. Consequently, an output file is given that gives information such as total fuel costs and the amount of MW heat produced by a specific boiler at a given hour. C3PO thereby calculates the amount of heat that is to be produced from each boiler, according to the demand of the buildings, and gives information about how the virtual storage is implemented.
In this master thesis, the focus has been to correctly model the customer flexibility in the grid. To do so the available flexibility must be quantified. This is done by creating an interpolation table, see Appendix A and table A.1. Interpolation is done to calculate the speed and the size of the storage at a specific outdoor temperature. How these values are calculated are described in the section 4.2.
4.2 Virtual Storage
Customer flexibility or short-term building thermal energy storage is actuality stored in the building envelope of the MFDs. However, in C3PO the flexibility is model as non- existent heat storage, referred to as virtual storage. The storage is temperature dependant and can be charged up or down depending on the flexibility in the DHS. See figure 4.1 for an illustration.
Figure 4.1: An illustration of the virtual storage
At a specific outdoor temperature, the speed and size of the virtual storage are given.
The speed describes the amount of power [MW] that can be charged or discharged from the virtual storage. It is calculated using the energy signature for the DHS. The energy signature is calculated using the consumption of the 50 largest MFDs in the specific DHS from the year 2018. The energy signature is presented below in figure 4.2.
Figure 4.2: Energy signature for 50 largest MFD in the DHS
The speed is calculated as the difference between the mean power for the total heating and the mean power for DHW. Thereby, the speed describes the power that is consumed by space heating. The heating power for DHW is not considered to not risk legionella [15]. Using the energy signature it is also calculated that the balance temperature for all
50 buildings is 16◦C. Temperatures higher than the balance temperature is in no need of space heating and instead only consume DHW. In those cases, customer flexibility is not effective and gives a speed of 0 MW.
The size of the virtual storage describes the amount of energy [MWh] that can be shifted during a flexibility period. The size of the virtual storage is the speed times the discharging time, t. The speed is already provided. The discharging time is the time it takes for the building to reach a predetermined indoor temperature variation. In this case 1◦C. See the equation below, where the discharging speed is isolated from equation 2.4.
t = ln(1 − ∆Ti/∆To) · (−τ )
When simulating it is possible to choose the flexibility period. This value is set to be a maximum of 1 day. The flexibility period decides between what period that the loads can be shifted. This value was validated using pilot results made in the chosen DHS. The pilot showed that loads can not be shifted however long in time, and must be limited by thermal dynamics. Since steering of the buildings comes from fluctuating the indoor temperature in building’, which creates short-term flexibility in the grid. This in turn limits the time in which loads can be used by the system. The pilot proved a flexibility period longer than one day is not realistic or possible in the specific DHS [27] [26].
4.2.1 Net Zero Energy
An important concept to explain is the Net Zero Energy (NZE) assumption. It describes the phenomenon that loads shifted must be equally as large. The assumption was there- fore that as much heat that was used to increase the temperature in a building will be added afterward or vice versa. Meaning that if the indoor temperature in a building is increased to “charge up” the virtual storage and therefore reduce the load before the peak, the same heat must be discharged from the storage at a later time. However, after the literature study, it became interesting to investigate the possibility to disregard this as- sumption. The reasoning behind that is that the indoor temperature decreases on average, compared to the reference case using no flexibility [10]. This is also interesting because it could be compared to values found in pilots results performed in the case study DHS. If internal gains are assumed in the energy calculations, less heat needs to be provided to the buildings. This phenomenon is described by equation 2.5 where a larger value Qgain reduces the amount of heat demand from the buildings Qheat. Not considering the NZE assumption is modeled as additional heat gains, such as solar gains in C3PO. This creates even further flexibility in the DHS because it creates an energy demand reduction, where heat is “saved” in the system. See figure 4.3 for a mathematical model of this concept.
Figure 4.3: Mathematical model of the virtual storage when NZE is rejected
Figure 4.3 above gives a mathematical model of the virtual storage for when the NZE is rejected. The storage can be discharged to 100 % but does not need to be fully recharged.
In scenarios where the NZE assumption is made, the heat storage is covered by a share of internal gains instead of 100 % produced heat. Let’s assume a share of 10% internal gains. This value describes the amount of heat that is gained from tenants’ activity and is a share compared to the total heat demand in the building during a short-time period.
Assuming that, it means that only 90 % of the energy that is to be recharged to the virtual storage is needed, instead of 100 %. Thereby, 10 % of the loads that are to be provided to the virtual storage are not needed due to internal gains. This creates even further flexibility in the DHS and reduces the heat demand from the buildings.
Scenarios assuming NZE and others for which the assumption is rejected will be simulated.
4.3 Key Assumptions
Assumptions made regarding the virtual storage are presented in table 4.1 below. These were validated using pilot results from the specific DHS.
Table 4.1: The values of the assumption assumed Indoor temperature variation, ∆Ti 1◦C
Time constant ,τ 100 h
Flexibility period 1 day
Flexible load 30 % of the buildings’ total energy consumption Charging speed Five times slower than the discharging speed
Internal gains 0-50 %
The indoor temperature describes the maximum temperature change that is allowed in the MFDs. It is set to be maximum 1◦C to not affect the thermal comfort as mentioned earlier. The time constant is set to 100 hours and it is a conservative value that describes all 50 MFDs since they are modeled as one. The flexibility period describes the time in which the loads can be shifted. It is set to be 1 day, meaning that the loads shifted on Monday morning can not be used on Tuesday for example. This value is also validated.
The same goes for the flexible load, charging speed, and internal gains. The flexible load describes the total amount of load that can be steered. This value is set to be 30 % of the buildings’ total energy consumption. 100 % is not used because it affects the secondary system of the DHS. The charging speed of the virtual storage is set to be five times slower than the discharging speed. Lastly, the internal gains are validated to be in a spectrum between 0 % and 50 % of the total heat demand in the buildings[27] [26].
All of the key assumptions are implemented for each simulated scenario of the DHS.
However, the last assumption regarding internal gains is varied between 0-50 %. The scenarios are presented in the next section.
4.4 Two Optimization Strategies
To calculate the KPIs two types of optimizations strategies are implemented in C3PO.
Firstly, a normal-operation optimization that caters to the reduction of the daily load variation and prioritizes optimization of the fuel costs. Secondly, a peak-operation opti- mization was made to investigate the potential of using customer flexibility to reduce the design load.
4.4.1 Normal-operation Optimization
C3PO is modeled to do an economic optimization and therefore does a normal-operation optimization in its default setting. Figure 4.4 below displays a heat production curve using the normal-operation optimization strategy.
Figure 4.4: Normal-operation optimization
4.4.2 Peak-operation Optimization
The peak-operation optimization is achieved by asserting an additional price-limitation of fuel cost at a specific peak power. To achieve the optimal reduction of peak power iterative guesses of peak power was made to estimate how low the peak power could be for a certain scenario. Evidently, with the limitation of still providing enough of the demanded heat load in the DHS. Figure 4.5 below displays a heat production curve using the optimization strategy.
Figure 4.5: Peak-operation optimization
4.5 Scenarios
The scenarios simulated are shown below in table 4.2 and 4.3. All cases are modeled for production in 2022 and therefore use predicted fuel prices for the mentioned year. The reason to why 2022 is chosen is because that is when new installations in of boilers are assumed to be in place. Moreover, distribution losses of 10 % are assumed [15]. The changing parameters between each scenario are the use of customer flexibility and energy reduction. All cases assume flexibility, an indoor temperature variation of 1◦C is assumed, and a flexibility period of 1 day.
Table 4.2: Definition of scenarios for normal-operation optimization
Scenario Number Scenario Name No Flexibility/ Flexibility Demand Reduction (NF/F) (0, 10, 30, 50) %
1 Baseline NF -
2 Normal Flexibility F 0
3 10% Demand Reduction F 10
4 30% Demand Reduction F 30
5 50% Demand Reduction F 50
Table 4.2 above shows that the Baseline scenario implements no flexibility, this scenario will be used as a reference scenario. Normal Flexibility is referred to as the scenario that implements customer flexibility in the DHS while assuming NZE. Meaning, no demand reduction of the buildings. Scenario 3-5 assume 10 %, 30 % and 50 % respectively demand reduction while implementing customer flexibility.
Table 4.3: Definition of scenarios for peak-operation optimization
Scenario Number Scenario Name No Flexibility/ Flexibility Demand Reduction (NF/F) (0, 10, 30, 50) %
6 EOT5 Baseline NF -
7 EOT5 Normal Flexibility F 0
EOT5 Baseline is a reference scenario for the peak-operation optimization scenarios and assumes no customer flexibility. The name of the scenario differs from scenario number 1 because a different set of data is used for these two scenarios. They are named EOT5 because these scenarios are used to investigate the effects of customer flexibility on the design load. Therefore, historical data for the DHS during the coldest year is used instead.
For the particular DHS, data for 2016 was used. The EOT5 Normal Flexibility scenario implements customer flexibility and assumes no demand reduction.
Results given from the simulation are presented in diagrams showing the production curve of the DHS in a chosen time period.
4.6 Calculating The KPI Values
In the results, chapter 5, the production curves for each scenario will be presented. More- over, their respective KPI values will be presented as well. In this section, information about the significance of each KPI and how these are calculated are provided.
4.6.1 Peak Power
The peak power is given in the unit MW, and is the maximum annual power at a specific hour used by the boilers in a particular DHS. The value varies from year to year because the heat production is temperature dependant. The peak power is found when the outside temperature is the lowest, i.e. the wintertime. Thereby higher peaks are found in the heat production to cover the space heating demand in the building. Calculating this value for the chosen scenario gives information about how much of the boiler capacity is used and can help dimension the DHS. It is calculated from the C3PO results file for each scenario by finding the maximum added power from all boilers in the DHS at a specific hour.
4.6.2 Peak Fuel Usage
The peak fuel usage is given in the unit MWh and is the amount of peak fuel consumed annually. In this case, it is referring to the bio-oil usage. This KPI value gives information about the peak production in the grid. The aim is to reduce this usage, by shifting loads to cheaper fuels. Therefore, the KPI value is a measure of how the implementation of customer flexibility affects the peak operation. The value is calculated by adding the heat produced from the bio-oil plants in the C3PO result file.
4.6.3 Produced Volume
Produced volume is a KPI that measures the total amount of heat produced in the grid annually, the unit is MWh. Here it is interesting to show the difference between the scenarios where internal gains are assumed compared to the scenarios that assume NZE.
This KPI is calculated by adding the total amount of MWh produced from all boilers in the simulated year.
4.6.4 Total Fuel Cost
For this KPI C3PO gives a total value in the result file for each scenario. These are provided in the unit SEK. Given the values for each simulation, the economic viability of each scenario can be compared.
4.6.5 Fuel Cost per MWh
The fuel cost per MWh KPI is calculated using the KPI value for the total fuel cost and the produced volume. By dividing the annual fuel cost by the annual produced volume in a scenario is calculated. The fuel cost per MWh is given in the unit SEK/MWh. This KPI is of interest because it shows how optimized the system is in terms of produced MWh heat.
4.6.6 Climate Footprint
The climate footprint is the first environmental KPI used in the analysis of the customer flexibility on the DHS. It is measured in the unit tCO2-equivalents. The higher the KPI value is for a specific scenario of the DHS, the higher the environmental impact is. The value is calculated by multiplying the key values in gCO2/kWh produced heat for a specific fuel [28]. The key values are presented in Appendix D in table D.1.
4.6.7 Primary Energy
The last KPI value is primary energy, measured in the unit MWh, and gives information about the amount of natural resources needed to produce heat. Raw resources are for example natural gas, waste, biomass, hydro, and wind [29]. The KPI value is calculated by multiplying the total amount of supplied energy to the DHS for a specific fuel with the primary energy factor for that fuel. These values are provided in table D.1 in the Appendix. The supplied energy is calculated by dividing the produced volume in a specific scenario with system efficiency, which considers losses in the delivery of the energy. The primary energy factor is a quota between the primary energy in [kWh] and the supplied energy in [kWh]. Thereby, using the factor the total primary energy is calculated for each scenario.
Chapter 5 Results
In this chapter, the simulation results from the modeling are presented. Moreover, all KPI results for the decided scenarios are given. The results are categorized between the two optimization strategies, normal-operation optimization, and peak-operation optimization.
The DHS is simulated for the year 2022 to investigate the future benefits of implementing customer flexibility. Fuel costs used are based on forecast fuel prices and 2022 boiler capacity. Moreover, it should be stated that the production from the different boilers is in a hierarchy based on their fuel prices. The cheaper the fuel, the lower priority. The cheapest fuel is biogas and the most expensive is biodiesel.
5.1 Normal-operation Optimization
The simulation results presented are those mentioned in section 4.5. The heat production curves for the scenarios are presented, given by C3PO. Consequently, the 7 KPIs are cal- culated based on the output data from the modeling. For the normal-operation scenarios, the heat curves are presented in the coldest 2-weeks of the year which are between the 21th February and 6th March 2022. Data used to model the energy demand is from 2018 but is scaled up to 2022.
5.1.1 Baseline
The first scenario presented is the baseline. The values calculated for this scenario are used as a reference for all normal-operation scenarios. Below in figure 5.1 the heat production curve for the scenario is shown.
Figure 5.1: Heat production curve for Baseline scenario
In this production curve it can be seen that no biodiesel is used, majority of the peak production is bio-oil. Moreover, the peak power, peak fuel usage, produced volume, fuel cost, climate footprint and primary energy usage of this scenario will be used to compare all normal-operation scenarios.
5.1.2 Normal Flexibility
The second scenario considers flexibility in the DHS. Below in figure 5.2 the heat produc- tion curve is presented.
Figure 5.2: Heat production curve for Normal Flexibility scenario
In the figure, it is visible that less bio-oil is used, compared to the baseline scenario which is represented by the dark blue line at the top. Therefore, it can be stated that loads have been shifted. The use of flexibility resulted in changes in the KPI values, compared to the baseline scenario. The KPI results are presented below in table 5.1.
Table 5.1: KPI values for Normal Flexibility scenario
KPI Normal Flexibility
Peak power [%] 1.9
Peak fuel usage [%] -11.1 Produced volume [%] 0.0
Total fuel cost [%] -0.9 Fuel cost per MWh[%] -0.9 Climate footprint [%] 1.2
Primary energy [%] 0.9
5.1.3 10 % Demand Reduction
In this scenario, the NZE assumption is not considered. Consequently, not only has heat load been shifted, but the overall produced volume reduces. Below in figure 5.3 the heat production curve for the scenario is presented.
Figure 5.3: Heat production curve for 10 % Demand Reduction scenario
The KPI results for the scenario are presented below in table 5.2.
Table 5.2: KPI values for 10 % Demand Reduction scenario
KPI 10 % Demand Reduction
Peak power [%] 1.9
Peak fuel usage [%] -12.4
Produced volume [%] -0.2
Total fuel cost [%] -1.3
Fuel cost per MWh[%] -1.1
Climate footprint [%] 1.1
Primary energy [%] 0.8
5.1.4 30 % Demand Reduction
Here, 30 % demand reduction is used instead of 10 %. Below in figure 5.4 the heat production curve for the scenario is presented.
Figure 5.4: Heat production curve for 30 % Demand Reduction scenario
Even less bio-oil is used as seen in the figure. The KPI results are presented below in table 5.3.
Table 5.3: KPI values for 30 % Demand Reduction scenario
KPI 30 % Demand Reduction
Peak power [%] 1.9
Peak fuel usage [%] -15.0
Produced volume [%] -0.6
Total fuel cost [%] -1.8
Fuel cost per MWh [%] -1.2
Climate footprint [%] 0.8
Primary energy [%] 0.5
5.1.5 50 % Demand Reduction
Lastly, another scenario being investigated is 50 % Demand Reduction. Below in figure 5.5 the heat production curve for the scenario is presented.
Figure 5.5: Heat production curve for 50 % Demand Reduction scenario
The KPI results are presented below in table 5.4.
Table 5.4: KPI values for 50 % Demand Reduction scenario
KPI 50 % Demand Reduction
Peak power [%] 1.9
Peak fuel usage [%] -16.3
Produced volume [%] -1.0
Total fuel cost [%] -2.4
Fuel cost per MWh[%] -1.4
Climate footprint [%] 0.2
Primary energy [%] 0.0
5.2 Peak-operation Optimization
There are two peak-operation scenarios; EOT5 Baseline and EOT5 Normal Flexibility.
The data was scaled up to 2022 heat demand as for the other scenarios. The coldest five days is assumed to be between the 15th January and the 19th January 2022. The key assumptions are the same as for the other scenarios, while 0 % was assumed for the EOT5 Normal Flexibility scenario. The flexibility period is set to a maximum of 1 day.
5.2.1 EOT5 Baseline
The baseline for the peak-operation optimization is used as a reference for the EOT5 Normal Flexibility scenario. The heat production curve for the reference scenario is presented below in figure 5.6.
Figure 5.6: Heat production curve for EOT5 Baseline scenario
5.2.2 EOT5 Normal Flexibility
In this scenario, the effects of flexibility on the peak load are to be investigated. The heat production curve is presented below in figure 5.7.
Figure 5.7: Heat production curve for EOT5 Normal Flexibility scenario
The KPI results for the scenario are presented below in table 5.5. A reduction of the peak load with 10.9 % is found.
Table 5.5: KPI values for EOT5 Normal Flexibility scenario
KPI EOT5 Normal Flexibility
Peak power [%] -10.9
Peak fuel usage [%] -26.0
Produced volume [%] 0.0
Total fuel cost [%] -3.6
Fuel cost per MWh[%] -3.7
Climate footprint [%] 1.0
Primary energy [%] 0.8
5.3 KPI Fulfillment
Below in table 5.6 all KPI results from the scenarios are presented. The normal-operation scenarios are compared to its baseline scenario and the peak normal flexibility is compared to its respective baseline from the peak-operation optimization.
Table 5.6: KPI fulfillment for all scenarios
KPI Normal 10% Demand 30% Demand 50% Demand EOT5 Normal Flexibility Reduction Reduction Reduction Flexibility
Peak power [%] 1.9 1.9 1.9 1.9 -10.9
Peak fuel usage [%] -11.1 -12.4 -15.0 -16.3 -26.0
Produced volume [%] 0.0 -0.2 -0.6 -1.0 0.0
Total fuel cost [%] -0.9 -1.3 -1.8 -2.4 -3.6
Fuel cost per MWh[%] -0.9 -1.1 -1.2 -1.4 -3.7
Climate footprint [%] 1.2 1.1 0.8 0.2 1.0
Primary energy [%] 0.9 0.8 0.5 0.0 0.8
From the table, it is shown that the normal-operation scenarios gave a small increase in the peak load. However, the peak-operation optimization gave a reduction of the peak load
with 10.9 %. Peak fuel usage decreased for all cases, which proves that flexibility reduces peak production. Moreover, the scenarios assuming NZE are confirmed to not affect the overall produced volume. Whereas, the scenarios that assume a demand reduction had a reduction in produced volume. The fuel costs reduced for all scenarios both in total and per MWh produced. The climate footprint and the primary energy also reduced for all scenarios that made use of customer flexibility.
