Energy flexible building through smart demand-side management and latent heat storage

1 Energy flexible building through smart demand-side management and latent heat storage Jesus Lizana a Daniel Friedrich b Renaldi Renaldi b Ricardo Chacartegui c a Departamento de Construcciones Arquitectónicas, Universidad de Sevilla, Avda. Reina Mercedes 2, 41012 Seville, Spain b School of Engineering, The University of Edinburgh, Edinburgh, EH9 3JL, UK c Departamento de Ingeniería Energética, Universidad de Sevilla, Camino de los Descubrimientos s/n, 41092 Seville, Spain Abstract One of the greatest challenges for long-term emissions reduction is the decarbonisation of heating and cooling due to the large scale, seasonal variation and distributed nature. Energy flexible buildings with electric heating, smart demand-side management and efficient thermal energy storage are one of the most promising strategies to deploy low-carbon technologies which can benefit the electricity system by reducing the need of reinforcing existing networks and their ability to use electricity in times of low demand and high supply. Combined with spot price contracts, in which the electricity tariff changes every half-hour depending on supply and demand, they can effectively reduce on-peak demand periods, achieve economic profits for end-users and retailers, and reduce the environmental impact of the electricity grid by operating in periods with lower CO2 emissions rate. To achieve these benefits, it is crucial to develop accurate models for energy flexible buildings as well as control strategies to optimise the complex system operation. This paper proposes a novel flexible energy building concept, based on smart control, high density latent heat storage and smart grids, able to predict the best operational strategy according to the environmental conditions, economic rates and expected occupancy patterns. The smart integration model, carried out in TRNSYS for a Scottish case study, solves a multi-criteria assessment based on future energy demand prediction (learning machine model supported by end-user’s predefined occupancy by Internet of Things, present and forecast weather data, and building load monitoring), electricity tariff evolution and building performance. The results show that end-user’s electricity bill savings of 20% are obtained and retailer’s associated electricity cost is reduced by 25%. In addition, despite an increase in final energy consumption of up to 8%, the environmental impact remains constant due to operation at times with lower CO2 emissions rate of the electricity generation. The developed tools enable the design of smart energy systems for energy flexible buildings which can have a large, positive impact on the building sector decarbonisation. Keywords: Thermal energy storage, Phase change materials, Low-carbon energy, Energy flexible buildings, Demand-side management, Demand response 2 Contents 1. Introduction....................................................................................................................................................................... 4 2. Description of research .................................................................................................................................................... 6 2.1. Definition of smart low-carbon heating solution in electricity network ....................................................................... 6 2.2. Scenarios for environmental and economic assessment .......................................................................................... 8 3. Methodology for numerical simulation ............................................................................................................................ 10 3.1. Building modelling ................................................................................................................................................... 12 3.2. Energy system modeling ......................................................................................................................................... 12 3.3. Conventional and smart demand-side management controller modelling .............................................................. 15 3.3.1. Conventional controller .................................................................................................................................... 15 3.3.2. Smart DSM controller towards flexible energy buildings .................................................................................. 15 3.5. Application example of smart DSM controller along a week with spot price contract.............................................. 19 4. Simulation results and discussions ................................................................................................................................ 21 4.1. Energy balance ....................................................................................................................................................... 21 4.2. Economic performance ........................................................................................................................................... 24 4.3. Environmental performance .................................................................................................................................... 25 5. Conclusions .................................................................................................................................................................... 26 Acknowledgments .............................................................................................................................................................. 27 References ......................................................................................................................................................................... 27 3 Nomenclature and abbreviations ASHP air source heat pump AWHP air-to-water heat pump 𝑐𝑝 specific heat, KJ/kg K COP coefficient of performance DH district heating DHW domestic hot water DR demand response DSM demand-side management E7 Economy 7 Fn correction factor according to demand prediction ℎ latent heat of fusion per unit mass, kJ/kg HDD heating degree days HP heat pump HVAC heating, ventilation and air-conditioning HX heat exchanger LT limit tariff, £/kWh IoT internet of things Pspecific buildings' specific heat loss rate PCM phase change material t time T temperature, ºC TL limit electricity tariff price TES thermal energy storage WWHP water-to-water heat pump Greek letters  thermal conductivity, W/m K 𝜌 density, kg/m3 Subscript l liquid s solid Superscripts aux auxiliary heater cond heat conduction hx immersed heat exchanger loss loss with the environment 4 1. Introduction The greenhouse gas emissions reduction targets set by many countries and agreed as part of the Paris agreement require almost the complete decarbonisation of the building sector. One of the greatest challenges to achieve these emissions reductions is decarbonising heating and cooling. Heating and cooling is responsible for around 50% of final energy consumption in the European Union [1]. While there has been significant progress in the decarbonisation of the heating and cooling sector, 75% of the current fuel consumption still comes from fossil fuels (nearly half from gas) [1]. The main challenge towards a low-carbon building sector consists of the large deployment of renewable energy sources along with most efficient generation technologies. Current policies enforcing decarbonisation in the electrical grids will result in an increasing electrification of building systems [2]. Thus, considering typical operating efficiencies and carbon footprint of heating technologies for buildings [3,4], different authors state that most promising solution for the future building sector would be a combination of TES solutions linked to efficient solar district heating (DH) networks and individual heat pumps in the remaining houses [5], involving the decarbonization of electricity networks [6]. This integration was previously identified by Lund el al. [7], Handy et al. [8] and Lizana et al. [9,10], who showed that the implementation of efficient and renewable DH systems, along with local efficient heat pumps (HPs), could play an important role in the successful deployment of a sustainable building sector. Heating, cooling and the electricity system can support each other in the effort to decarbonise. It is essential to recognise the links between them and exploit synergies. Part of the challenge involves decarbonising electricity supplied from the national grid, and then electrifying heating and cooling, by replacing most of the oil and gas boilers used to heat buildings today with technologies such as heat pumps which run on electricity [6]. However, the electrification of heat is expected to be a difficult task. It could cause a significant increase of electricity peak demand, which could have adverse consequences on the electricity system, in particular on the low-voltage distribution networks that deliver power from the substations along cables down residential streets [11]. This might be worsened by the future deployment of electric vehicles which might lead to high uncoordinated plug-in concentrations, significantly increasing on-peak electricity demand [12–14]. The cost of having to reinforce existing electricity networks to accommodate these new loads, heat pumps and other low-carbon technologies could be very considerable and therefore it is important to make best use of existing network assets, and to ensure that any reinforcement is based on an accurate assessment of need. Energy flexible buildings through smart demand-side management (DSM) or smart demand response (DR) using efficient energy storage are one of the most promising options to deploy low-carbon technologies in the electricity networks without the need of reinforcing existing networks [15,16]. Most common DSM strategies consist of the shifting of electrical loads (or demands) from on-peak to off-peak electricity tariff hours, and the improvement of energy performance of systems through the implementation of operational strategies by avoiding partial load operation or operating at other sub-optimal times [5]. Different studies based on DSM strategies have demonstrated their high economic savings due to the advantage of electricity prices along off-peak tariff hours. Main DSM strategies are based on the implementation of thermal energy storage (TES) units using sensible, latent or thermochemical heat storage, the use of passive thermal storage of the building to shift the electricity demand, and the deployment of distributed electrical storage using electrical batteries at the building level. As DSM applications based on sensible storage, Renaldi et al. [17,18] evaluated the economic performance of an air source heat pump (ASHP) linked to a water tank to shift electrical load from on-peak to off-peak electricity tariff periods. In [17] the results showed that the integration of a storage tank of 180l was able to shift approximately 13% of on-peak demand to off-peak time, achieving an operational cost reduction of 21.8% in comparison with using a conventional ASHP without TES. In [18] they assessed a storage tank of 300l, achieving an annual operation cost reduction between 26.6% and 29%. They demonstrated that annual operational cost decreases as storage capacity is increased. Arteconi et al. [19] evaluated an air-to-water heat pump (AWHP) linked with a water tank for heating purpose through a DSM strategy to shift the onpeak demand to off-peak times. Two different heating distribution systems were assessed: underfloor heating and radiators. The results showed that the underfloor heating system (always on), operating by a DSM strategy, reduced the electricity 5 bill by around -9%. However, annual energy consumption was slightly higher (around +3-4%). For radiators, operating 8 hours per day, the DSM strategy did not reduce the electricity bill (around +4%) and the annual energy consumption was higher (around +8-9%). They concluded that using TES and a DSM strategy no reduction of energy consumption is achieved, but if a time-of-use tariff is available and an effective DSM strategy is implemented, the electricity bill could be cut down. In other research, Arteconi et al. [20] evaluated water-to-water heat pumps (WWHPs) coupled with a water tank for cooling purposes through a DSM strategy (TES charged by the heat pumps during off-peak hours). The integration saved 54% of energy cost and over 42% of energy compared with normal HPs operation. Yan et al. [21] evaluated a novel air-conditioning system with proactive demand control for daily load shifting and real-time power balance. It consists of chillers and a chilled water storage tank. The aim was to enable more flexibility by changing its electricity usage pattern. Compared with conventional air-conditioning system, the chiller capacity was reduced by 24% and the total energy cost was reduced by 29.7%. The use of existing building fire service water tanks as chilled water storage for power demand limiting was assessed by Yan et al. [22]. Two optimal demand limiting control strategies were proposed (time-based control and demand-based control) for maximizing the monthly peak demand reduction of buildings with regular and variable peak occurring time. The results showed that substantial cost savings can be achieved in energy retrofitting through this application with short payback periods (0.7–2.6 years). As DSM applications based on latent heat storage and thermochemical storage, Cui et al. [23] evaluated the performance of a DSM strategy in commercial buildings based on a chiller and an air handling unit, integrated with cold phase change material (PCM) storage. The results showed that cool TES implementation for shifting peak demand achieved significant annual net cost savings, as well as providing longer duration and larger capacity for demand response. This resulted in a payback period of 8-9 years. Finck et al. [24] quantified the demand flexibility of an office building using different storage tanks. They identified that using a water tank, a phase change material tank or a thermochemical material tank integrated for heating needs in buildings through an optimal control can supply high energy demand and provide power flexibility along with economic savings. As DSM application based on building thermal mass storage, Romero et al. [25] analysed the performance of a HP and PV panels integrated with different dynamic electricity pricing strategies in a German case study using TRNSYS. A smart controller was implemented to automatically manage the activation of the HP in response to day-ahead electricity pricing changes with the aim of cost savings and reducing the overload of the grid during peak periods. The energy was stored in the thermal mass of the building envelope, offering a passive thermal storage to shift the electricity demand. The heat pump is connected directly to the underfloor heating manifold, and the heating tank is located on the return flow. This uncommon configuration had the advantage of focussing on the thermal storage capacity of the building, not on the thermal tank. Cost savings up to 25% were achieved for scenarios where HP operation was limited to lower electricity price periods. As DSM application through distributed electrical storage, Sandoval et al. [2] evaluated the electricity demand flexibility of using different electric batteries at the residential levels. They highlighted the need of the active role of the building as a coordinator between the electricity utility and the final consumers and observed that the battery component capital cost is more than twice as large as the PV capital cost. Many reported case studies show that small-scale TES units are feasible and technically attractive solutions. The performance of low-carbon energy technologies through TES implementation is clearly enhanced following DSM strategies. Furthermore, higher volumetric TES density with reduced heat losses will allow greater benefits. However, previous DSM studies based on night consumption periods using fixed off-peak electricity tariff cannot effectively represent the situation in electricity grids with ever increasing renewable generation capacities. For example, in the UK the net demand, i.e. total demand minus renewable generation, is sometimes higher during the night due to high solar PV output [26]. As a step forwards, the integration of HPs and high-density TES combined with smart DSM strategies based on real-time and predictive electricity data can efficiently support the transition towards a low-carbon energy sector. This smart integration is able to operate at optimal times from the point of view of electricity grid and consumers, getting high environmental and economic benefits. The building sector can play an important role in the final electricity demand profile, 6 and can have significant impact on the existing and projected electrical system infrastructure. Their ability to control the timing and magnitude of the aggregate electricity consumption acquires critical relevance. Thus, with the aim of assessing the potential benefits of smart energy flexible buildings, this research evaluates a novel application of smart DSM based on the integration of a HP, an efficient and high density latent heat thermal energy storage (LHTES) unit and a smart controller. The smartest operational strategy is managed by a controller which solves a multi-criteria assessment based on future energy demand prediction (learning machine model supported by end-user’s predefined occupancy, present and forecast weather data, and building load monitoring), electricity tariff evolution and building performance. The paper is structured as follows. First, the smart low-carbon heating application is further described, and the case study and scenarios are characterised. Afterwards, the methodology for numerical simulation, which was carried out in TRNSYS, is detailed. It was divided into three groups: building modelling, energy system modelling and controller. Finally, energy balance, and economic and environmental performance are presented and discussed. 2. Description of research 2.1. Definition of smart low-carbon heating solution in electricity network The smart integration of DSM strategies in domestic buildings has the potential to flatten the electricity demand profile throughout the day as illustrated in Fig. 1 Figure 1. Future electricity grid performance through smart flexible energy buildings. A smart system able to predict the best available operating schedule according to the environmental conditions, renewable generation, economic rates and expected occupancy patterns can generate an efficient building energy system with benefits for the consumer, the electricity grid and the environment. It would require the support of efficient TES systems, developed and conceived for full integration in buildings. This manuscript assesses the proposed smart and novel lowcarbon integration which is illustrated in Fig. 2. 7 Figure 2. Smart heating and DHW integration towards flexible energy buildings. It consists of a heat pump providing heat for heating and domestic hot water (DHW), a latent thermal energy storage unit and a smart DSM controller. The controller decides the most favorable way of energy management, using a latent heat storage unit as support system, which enables the heat pump to operate at optimal times, with the aim of reducing the energy bill. For example, the heat pump is operated when exterior conditions are better for increasing the coefficient of performance (COP) of the system, or when the electricity tariff is lower due to low-energy demand and/or high renewable generation. The smart DSM controller assesses the future energy demand according to end-user’s predefined occupancy by a phone app based on Internet of Things (IoT), weather data (present and forecast) and building load monitoring. Thus, the system stops the charging period if no future energy demand is required, or if stored energy is enough to supply energy needs. In addition, electricity tariff evolution is considered to decide the most favorable economic period for electricity consumption. This integration linked to novel spot price contracts, in which electricity tariff that retailers offer to consumers is based on electricity spot price, changing every half-hour depending on supply and demand, leads to flexible energy buildings able to get high economic and environmental benefits by operating at low tariff periods, associated to low CO2 emissions rates and low electricity demand in electricity grid. Fig. 3 shows the technical and economic structure of the electricity market [27]. 8 Figure 3. Technical and economic structure of electricity market. The electricity price that retailers pay when they buy electricity from the wholesale market is electricity spot price. It changes every half-hour and can vary quite dramatically. Most electricity retailers charge customers a fixed electricity tariff based on a flat rate or two rates (such as Standard or Economy 7 electricity tariffs). This means consumers pay the same rate or rates for the electricity they use regardless when they use it, and regardless of the spot price their retailer paid. The flat rate contains the electricity spot price, final retailer’ profits, along with other additional costs, such as transmission and distribution costs, taxes or supplier surcharge. According to Nord Pool data [27] for the UK grid for different years, the consumer cost is between 3.1 and 3.5 times higher than the average annual electricity spot price. On the other hand, a novel electricity tariff contact is being proposed for retailers in some countries, called spot price contract, in which electricity tariff is based on electricity spot price. Daily electricity trading is mostly driven by the day-ahead market. The buyer (in this case the retailer) needs to assess how much energy it will need to meet demand the following day, and how much it is willing to pay for this volume, hour by hour. Then, the intraday market supplements the day-ahead market and helps to secure the necessary balance between supply and demand in the power market. Thus, through a large deployment of this new concept, retailers could reduce the associated cost of buying electricity in the day-ahead and intraday market by moving end-user’s consumption to off-peak periods. This enables a more competitive electricity cost for end-users as well as a better utilisation of the grid and generation infrastructure. 2.2. Scenarios for environmental and economic assessment Fig. 4 shows a conventional assembly schematic of an air-to-water heat pump for space heating and hot water. It has been used as reference scenario for evaluating the economic profitability and environmental performance of the proposed lowcarbon heating solution. This conventional heating and DHW system operates with standard electricity tariff, which cost rate is defined in Table 1. 9 Figure 4. Conventional heating and DHW system. Table 1. Electricity tariffs Tariff On-peak (£/kWh) Off-peak (£/kWh) Standard 0.144 0.144 Economy 7 0.1747 0.0765 Electricity spot price contract Changeable every half-hour depending on supply and demand Along with the reference scenario, two smart DSM integrations are assessed, following the model illustrated in Fig. 2: one based on Economy 7 (E7) tariff and another based on spot price contract. E7 is a two-rate tariff with off-peak duration from 00.00 to 07.00. Electricity spot price contract is a changeable tariff per hour, depending on electricity spot price. A new detached 2-storey house in Lochearnhead, Scotland, has been selected for assessing the proposed DSM strategies. Scottish climate is characterised by warm summers, and cold and wet winters. Heating is the main energy demand in Scottish energy sector, responsible for 53% of energy consumption, followed by transport (25%) and electricity (22%) [28]. Main primary heating fuel is gas, covering 79% of heating needs, followed by electricity (12%) and oil (7%) [28]. For an average household, hot water and space heating represent approximately 87% of total energy use [29]. Table 2 shows the main fabric parameters and compares the values with limiting standards for the properties of the fabric elements of the building, according to UK Building Regulations [30]. This case study consists of a highly-insulated, lowenergy demand house. The characterisation of the building energy system is reported in Table 3. Case study is equipped with a heat recovery ventilation system with an effective recovery rate of 93% and an efficient air-to-water HP with a COP of 3.13, which provide heat for heating and hot water. Table 2. Characterisation of fabric parameters of the case study. Element Definition Characterisation Limiting values Window Triple glazing system with wooden frame U-value: 1.00 W/m2 K U-value: 2.00 W/m2 K Facade Cement-bonded wood fibre blocks containing insulating foam and concrete, and plywood U-value: 0.16 W/m2 K U-value: 0.30 W/m2 K Roof Soil (green roof), insulation and plywood U-value: 0.23 W/m2 K U-value: 0.20 W/m2 K Floor Floating floor concrete slab, insulation and concrete slab foundation. U-value: 0.25 W/m2 K U-value: 0.25 W/m2 K 16 Future heating demand prediction (𝐹𝑢𝑡𝑢𝑟𝑒 𝑄ℎ𝑒𝑎𝑡𝑖𝑛𝑔) in kWh is calculated for the next 24 hours based on an iterative self-learning procedure (or machine learning) through Heating Degree Days (HDD). HDD is a common and simple metric that allows quantifying the amount of heating or cooling that buildings in a specific location need over a certain period. It is calculated with a base temperature of 20ºC. Given the number of HDD, to calculate the energy required by a building, heat needs to be provided to the building at the rate at which it is being lost to the environment. This value is almost constant over a short period of time for each building, and is defined as buildings' specific heat loss rate (Pspecific), generally given in Watts per Kelvin (W/K). So, final future energy need is calculated according to Eq. 2. 𝐹𝑢𝑡𝑢𝑟𝑒 𝑄ℎ𝑒𝑎𝑡𝑖𝑛𝑔,𝑡=𝑃𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐,𝑡· (∑𝐻𝐷𝐷𝑓𝑢𝑡𝑢𝑟𝑒 𝑑𝑎𝑦 6 𝑑𝑎𝑦 1 /6)·∫ 𝑂𝑐𝑐𝑢𝑝𝑓𝑢𝑡𝑢𝑟𝑒 𝑡+96 𝑡4·1000⁄ (kWh) Eq. 2 Where 𝐻𝐷𝐷𝑓𝑢𝑡𝑢𝑟𝑒 is the average value of Heating Degree Days for future 6 days (144 hours) and it is estimated by means of weather forecast data. In the simulation model, this value is obtained from the weather data file. 𝑃𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐,𝑡 is the building’s specific heat loss rate, and is calculated according to past building performance of last 6 days (144 hours), according to Eq. 3, as a function of energy consumed, hours occupied and past weather data. When a shorter performance assessment period is considered, the results are not as representative due to the occasional peak temperature oscillation. The predictive model works better with an average value along a reasonable period. On the other hand, 𝑂𝑐𝑐𝑢𝑝𝑓𝑢𝑡𝑢𝑟𝑒 is the future occupancy hours along future 4 days since the storage unit is able to provide energy for 4 days along low demand periods. Following reported criteria, the predicted heating demand ranges from 90% to 110% of the actual heating demand. 𝑃𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐,𝑡= [ ∫ 𝑄𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 𝑡 𝑡−144 ·1000] [ ∫ 𝑂𝑐𝑐𝑢𝑝ℎ𝑜𝑢𝑟𝑠 𝑡 𝑡−144 ·(∑ 𝐻𝐷𝐷 𝑑𝑎𝑦−1 𝑑𝑎𝑦−6 /6)] ⁄ (W/K) Eq. 3 Present and future electricity tariff values. The smart DSM controller has to be able to decide when electricity should be consumed as a function of present and future electricity tariff evolution. Two types of electricity tariffs are considered for DSM strategies: Economy (E7), and Electricity spot price contract. They are summarised and compared with Standard electricity tariff, which was used for reference scenario assessment, in Table 1. Standard is a one rate tariff for the entire day. E7 is a two rate tariffs with off-peak duration of 7 hours, from 00.00 to 07:00. Electricity spot prices change every half-hour and can vary quite dramatically depending on supply and demand. It is the price retailers pay when they buy electricity from the wholesale market. It means that consumers will pay a variable rate according to the spot price that the retailer pays. Electricity tariff rates (Standard and Economy 7) are compared with electricity spot prices in Fig. 9. The average value of spot price throughout the year is 0.04 £/kWh, with minimum and maximum values of 0.007 and 0.99 £/kWh, respectively. Fig. 9 shows that the maximum spot price profile has two high cost or on-peak periods from 9am to 11am and from 4pm to 9pm. 17 Figure 9. Selected electricity tariff values per hour calculated along a year. Buying electricity through spot price contract has advantages. When spot prices are low, you could pay substantially less for electricity than you would on a traditional retail contract where the price is the same all the time. But, the same is true in reverse. If spot prices increase above the flat rates charged by retailers, you may pay substantially more than you would on a traditional retail contract. As the electricity price changes over time, the smart DSM controller divides the electricity price rates into 3 ranges (A, B and C), in which it implements a priority order of operation modes. Tariff period A consists of a lowest cost range, tariff period B is a medium cost range, and tariff period C involves the high cost range. So, the system supplies or stores as much energy as possible at lower electricity price periods, avoiding consumption at on-peak periods with high electricity prices. Electricity price ranges (A, B and C) are limited by two limit prices over time (t), LT1 and LT2, which are assessed according to Eq. 4. Tariff period A – Low cost: £/kWh𝑡  𝑇𝐿1𝑡=(2·∑𝑇𝑎𝑟𝑖𝑓𝑓 𝑡+24 𝑡24⁄ +7· min 𝑡 → 𝑡+12(𝑇𝑎𝑟𝑖𝑓𝑓)·𝐹𝑛)10⁄ Tariff period B – Medium cost: 𝑇𝐿1𝑡<£/kWh𝑡  𝑇𝐿2𝑡=∑𝑇𝑎𝑟𝑖𝑓𝑓 𝑡+24 𝑡24⁄ Tariff period C – High cost: £/kWh𝑡>𝑇𝐿2𝑡 Eq. 4 TL1 is calculated as a fraction between the average electricity price in the next 24 hours (from day-ahead prices) and the minimum electricity price in the next 12 hours. This fraction ensures a limit electricity cost covering an adequate daily period to supply required energy with the lowest electricity tariff. As the required daily time of electricity consumption oscillates between winter (5-8h) and summer (1-3h due to no heating demand), a correction factor is introduced (𝐹𝑛). It adjusts the 18 lower tariff period A with the aim of consuming electricity in the lowest off-peak period. 𝐹𝑛 oscillates between 1.15 and 1.25 as a simple linear regression based on future heating demand prediction (kWh). TL2 is calculated as the average electricity price throughout future 24 hours. The smart DSM controller considers all reported variables to decide the most favourable procedure of energy management. Operation strategies are illustrated in Fig. 10. They have been grouped in three groups: hot water supply, indirect heating supply and direct heating mode. Figure 10. Operating modes of smart heating and DHW system. 19 Hot water supply is divided into two operation modes, using HP linked to the water tank (mode 1a) or using electric heater placed at the bottom of the tank (mode 1b). Indirect heating supply is divided into two operating steps, charging LHTES unit and discharging LHTES unit for underfloor heating. Direct heating mode consist of HP directly running with underfloor heating system. As each operating mode has different operating temperatures, they run separately. For example, underfloor heating system operates with water flow temperature close to 35ºC and hot water system operates with a setpoint temperature of 55ºC. If we provide water flow at 55ºC to the underfloor heating system, it could be damaged. Programming of operational strategies and priority of operating modes are based on three steps which are summarized in Table 5. First step consist of the electricity tariff range in which energy demand is produced, second step refers to the priority of operating strategies implemented in the controller, and finally the compliance conditions for each operating mode according to parameters previously defined. The smart DSM controller update each parameter along each simulation time step (5minutes), changing operation mode according to operational strategies implemented. Table 5. Linear operating structure of smart DSM strategies. Step 1: Electricity tariff range Step 2: Priority of operation modes Main compliance conditions £/kWh ≤ LT1 Mode 1a. HP & Water tank -Short dead band of 40-55ºC Mode 1b. Auxiliary water heater -At regular intervals after operation mode 1a Mode 3. HP & Underfloor heating -Heating set-point of 17 or 21ºC -Occupancy and operating schedules Mode 2a. HP & LHTES unit -DHW tank dead band of 45-50ºC -Future heating demand > 20kWh -LHTES level < 0.75 · Future heating demand -Future 24h occupancy profile > 5 hours LT1 < £/kWh ≤ LT2 Mode 1a. HP & Water tank -Large DHW tank dead band of 35-55ºC Mode 3. HP & Underfloor heating -Heating set-point of 17 or 21ºC -Occupancy and operating schedules Mode 2a. HP & LHTES unit -Dead band of 45-50ºC -Future heating demand > 20kWh -LHTES level < 0.75 · Future heating demand -LHTES level < 8kWh LT2 < £/kWh Mode 1a. HP & Water tank -Large DHW tank dead band of 35-55ºC Mode 2b. LHTES unit & Underfloor heating -Heating set-point of 17 or 21ºC -Occupancy and operating schedules -LHTES level > 3.5 kWh Mode 3. HP & Underfloor heating -Heating set-point of 17ºC -Occupancy and operating schedules 3.5. Application example of smart DSM controller along a week with spot price contract. As an example of smart DSM operation, Fig. 11 illustrates the main operating parameters along a week in the building scenario using electricity spot price contract. Fig. 11 is divided into four sections which illustrate the occupancy and operating rate (%); the energy balance of LHTES unit (kWh), water tank (kWh) and future heating demand predication (kWh); the power consumption of HP (kW) and electric water heater (kW); and day-ahead electricity prices at real time, ranged by two limit prices over time, LT1 and LT2. Different marks have been placed to show how the operational controller responds to different parameter changes: -Mark 1: Building is occupied (Fig. 11A) but no energy input is required. -Mark 2: The electricity spot price is below LT1 (Fig. 11D), the HP is run to keep the DHW tank in the short temperature DHW dead band, using off-peak spot prices. Storage level of water tank increases up to approximately 8kWh (Fig. 11B). -Mark 3: Building is occupied (Fig. 11A) and electricity spot price is over LT2 (Fig. 11D), so LHTES unit provide heating supply to the underfloor heating system, discharging LHTES level (Fig. 11B). 20 -Mark 4: As electricity spot price is below LT1 (Fig. 11D), and future heating demand is close to 30kWh (Fig. 11B), HP charges LHTES unit, increasing storage level up to 23 kWh. -Mark 5: Same as mark 3. -Mark 6: After charging water tank (Fig. 11B) using off-peak spot prices (Fig. 11D), electric heater starts running (Fig 11C) to increase the water temperature above 60°C to inhibit legionella bacterium growth. -Mark 7: Heating is directly provided by HP to the underfloor heating system. -Mark 8: The electricity spot price was below LT1 (Fig. 11D), and the HP was run to keep the DHW tank in the short temperature DHW dead band. Then, electricity spot price increases over the limit L1 value, and HP stops. -Mark 9: As electricity spot price is below LT1 (Fig. 11D) and no energy is required for hot water and heating, HP charges LHTES unit up to future heating need predicted, which is approximately 20kWh (Fig. 11B). Figure 11. Visualisation of smart operational strategies along a week with spot price contract. 21 4. Simulation results and discussions Energy balance, economic performance from the point of view of different agents (end-user’s and retailers) and environmental impact of three scenarios (conventional operating mode, smart DSM operating mode with E7 tariff, and smart DSM operating mode with spot price contract) are assessed and discussed in this section. 4.1. Energy balance Energy performance for the three scenarios was calculated using the simulation output with a time resolution of 5 minutes. Table 6 shows the results of monthly energy consumption achieved for each operating strategy. The energy consumption is divided into electricity consumed by HP, and electricity consumed by electric heater placed in water tank to achieve 60ºC once a week. Table 6. Electricity consumption balance of each operating strategy. Conventional mode Smart DSM operating mode (E7 tariff) Smart DSM operating mode (Spot price contract) Months HP (kWh) Electric heater (kWh) HP (kWh) Electric heater (kWh) HP (kWh) Electric heater (kWh) January 429.9 9.0 507.6 6.2 474.4 5.4 February 408.9 7.7 475.9 6.2 461.2 5.9 March 408.9 7.2 487.1 6.0 460.1 6.0 April 308.8 8.3 407.6 5.7 347.7 6.5 May 152.9 4.2 184.7 6.5 154.7 5.1 June 69.2 6.0 75.7 4.8 64.7 3.3 July 59.3 5.7 65.7 4.4 58.9 3.5 August 57.1 6.5 60.2 4.4 50.4 5.7 September 198.2 7.7 230.8 4.8 212.3 5.0 October 312.4 8.9 377.6 6.3 349.5 6.9 November 375.6 8.4 417.0 5.6 364.4 6.0 December 399.2 7.2 486.1 5.6 449.3 6.3 Total 3180.4 86.6 3775.9 66.2 3447.6 65.4 The electricity consumption of the reference case (conventional operating mode) is 3266.9 kWh, considering heat pump and electric heater in DHW tank. Following the DSM strategies previously defined in methodology, electricity consumption increases by 18% through smart DSM operating mode with E7 tariff, and by 8% in DSM operating mode with spot prices. The E7 tariff strategy and the spot price contract strategy shift the heat consumption by 40% and 29%, respectively. Thus, higher energy shifting is leading to higher final electricity consumption. In addition, for the specific case of DSM operating mode, the temperature set-point for the smart DSM controller along the daytime was considered as 17°C during the onpeak electricity tariff prices with the aim of shifting consumption towards lower electricity tariff period. The percentage of time along the year in which the controller follows this operating strategy is 4%. It is a short period of time in which the enduser avoids high electricity cost by enlarging higher the set-point temperature band. Main reasons of the increasing energy consumption are related to heat exchanger losses (charging and discharging LHTES unit) and storage losses. They are characterised in Fig. 12, which illustrates the Sankey diagrams of the reference scenario (conventional operating mode) and two smart DSM scenarios assessed. DHW tank losses are similar in all three scenarios, ranging from 28 to 30% of heat supplied from HP for DHW. LHTES loss ratio for both smart DSM scenarios are lower than DHW tank due to more efficient storage unit with higher insulation, being situated between 5 and 8% of heat supplied from HP for indirect heating. Heat exchange losses of the conventional operating mode are 2-3% of total heat supply from HP. However, heat exchange losses of both smart DSM operating modes are higher, around 4-5%, which are associated to higher number of charging and discharging processes. 22 Figure 12. Sankey diagrams of the three different scenarios Fig. 13 illustrates electricity consumption per hour over the whole year for different smart operational strategies and compared with the reference operating mode (conventional heating system). 23 Figure 13. Electricity consumption accumulated per hour along a year for different Smart operational strategies. Following the smart DSM strategy with E7 tariff (Fig. 13A), electricity consumption is shifted from on-peak period (from 7:00 to 0:00) to off-peak period along the night. Using the smart DSM strategy with an electricity spot price contract (Fig. 13.B), electricity consumption is shifted from two on-peak periods previously identified in Fig. 9 (from 8:00h to 11:00h and from 16:00h to 10:00h) to off-peak periods with lower electricity spot prices along the day (along the night and from 12:00h to 15:00h). Despite spot prices changing every half-hour, daily spot price profiles are very similar. 24 4.2. Economic performance The economic performance of two smart DSM strategies are assessed from the point of view of end-users and retailers and compared with the reference scenario (conventional operating mode). Smart demand shifting can reduce the electricity bill of end-users due to lower electricity tariff periods and provides electricity flexibility when electricity spot prices are higher to achieve lower electricity cost for retailers. Table 7 summarizes final electricity cost and consumption of end-users and retailers for each assessed scenario. Retailer’s electricity cost provides the associated cost that retailer pays for electricity according to spot price (£/kWh). End-user’s electricity bill shows the cost that the consumer pays for electricity according to electricity tariff (Standard, Economy 7 or Electricity spot price contract), and end-user’s electricity consumption reports the annual final energy consumption per operating mode. Table 7. Economic performance of different operating strategies. Conventional mode (Standard tariff) DSM operating mode (E7 tariff) DSM operating mode (Spot price contract) Retailer’s electricity cost £ 166.82 (100%) £ 147.53 (88%) £ 125.19 (75%) End-user’s electricity bill £ 470.43 (100%) £ 452.61 (96%) £ 375.57 a (80%) End-user’s electricity consumption 3266.9 kWh (100%) 3842.0 kWh (118%) 3513.0 kWh (108%) a Final electricity bill using spot price contract was calculated by multiplying the electricity spot price by three. This value was fixed considering that the average value of spot price throughout the year is 0.04 £/kWh. From the point of view of retailer’s electricity cost, the smart DSM control reduces energy cost by 12% in E7 tariff scenario and by 25% in spot price contract scenario. These results show that electricity spot price contracts together with smart flexible energy buildings can lead to a promising smart management of energy sources, with high economic profits and reducing on-peak demand periods in electricity network. From the point of view of end-user’s electricity bill, the smart DSM control reduces the energy bill by 4% in E7 tariff scenario and by 20% in spot price contract scenario. Thus, despite an increase in energy consumption by 18% and 8% in both scenarios, respectively, economic savings are achieved. However, we should consider that annual economic saving of the best scenario, associated to the electricity spot price contract, is approximately £95, which means that the investment cost of smart integration along with LHTES unit should be lower than £950 to achieve a payback period of approximately 10 years. Along with the economic profits for retailers and end-users, there are additional benefits such as better utilisation of the electricity grid and the renewable resource, reduced need to curtail renewable generation and no requirement of upgrading the grid to support the installation of new electricity heating systems. 25 4.3. Environmental performance Energy flexible buildings not only allow operating when electricity cost is lower, but also can reduce the environmental impact of the building sector since electricity is consumed when low-carbon generation technologies are providing most of the energy. With the aim of calculating the environmental impact of each operational strategy, hourly CO2 emissions of the electricity grid were calculated considering the CO2 emissions rate of UK network according to grid data [48]. The CO2 emissions rate is the ratio between the total amount of CO2 emitted divided by the total electricity generated in each hour. Here, the electric energy generated by each generation technology from [48] is multiplied with the amount of carbon emitted by this technology for each unit of electricity generated from [49]. This value is summed for all generation technologies to get the total amount of CO2 emitted. Thus, the final CO2 emissions associated to each scenario are calculated by multiplying the hourly CO2 emissions rate with the hourly electricity consumption along the year. Table 8 shows the results of environmental impact for the three scenarios. Table 8. Environmental performance of different operating strategies. Conventional mode DSM operating mode (E7 tariff) DSM operating mode (Spot prices) End-user’s energy consumption 3266.9 kWh (100%) 3842.0 kWh (118%) 3513.0 kWh (108%) CO2 emissions 1015.84 kgCO2 (100%) 1078.37 kgCO2 (106%) 1028.71 kgCO2 (101%) The results show that despite increasing energy consumption for energy shifting from on-peak electricity periods to offpeak periods, the environmental impact is almost constant for DSM operating mode using spot prices, and only 6% higher for DSM operating mode using E7 tariff. The latter shows that in the changing electricity system the off-peak and lowcarbon periods are not confined to the night-time anymore and that the spot price contract is better at utilizing low-carbon, renewable electricity sources. The majority of producers that sell their electricity on the spot market, via day-ahead auctions, must send their sale offers (for the amount of electricity they are prepared to deliver at various prices during the 24 hours of the following day) to the power exchange by 12 noon on the day before the power is delivered to the grid. At the same time, electricity retailers must send their purchase orders (corresponding to the amount of electricity they believe customers will consume during the 24 hours of the following day), and the amount they are willing to pay. During off-peak periods, energy is characterised by cheaper and low-carbon operating technologies, such as renewable energy generation plants (hydro, wind or solar) or nuclear plants. They provide better costs for retailers’ demands. On the other hand, on-peak periods consist of more expensive electricity because it is necessary to operate the most expensive production plants to supply required demand, which are characterised by higher CO2 emission ratio per kilowatt hour. Thus, shifting consumption toward off-peak periods leads to environmental benefits due to the use of low-carbon energy sources and a better utilization of electricity grid and generation infrastructure. It is also expected that the environmental impact of energy flexible buildings will decrease with the further deployment of non-dispatchable, low-carbon electricity generation.