Delivering Smart Energy Systems

Transcription

1 Delivering Smart Energy Systems

2 Forewords Frong Li The UK reached a milestone on the 20th April 2017, the first ever working day in Britain withot coal since the indstrial revoltion!. The key challenge now is to extend the one day to 365 days a year. This reqires not only aggressive technical developments and deployments, bt critically, market arrangements and innovative bsiness models that deliver excellent vale to energy cstomers and bsinesses. I believe ebay and Airbnb are the models we shold be inspired by. This sharing economy approach has the potential to nlock vale in or energy systems, particlarly at the commnity level. Throgh high-speed ICT, disrptive bsiness models and horizontal energy markets, a sharing economy wold deliver affordable renewable-based energy prodcts, provide flexible and cheap network access by exploiting nder-tilised resorces sch as, spare capacity in or network, and storage, demand response and power electronics at cstomers premises. At the University of Bath, we condct research that offers holistic soltions towards a smart-sharing energy system, from local energy market designs, innovative bsiness models, flexibility enablers to big data analytics. Or key aim is to spport the indstry to transit from a carbon friendly high-asset-low-intelligent system to a renewable friendly low-asset-high intelligent smart system. We develop fndamentally new markets to directly trade renewable based energy prodcts. We develop shared network access concept, allowing third party network operators to provide fast and cheap network access for flexible demand. We design and optimise local energy management system that can maximise retrns from both horizontal and vertical markets. We are delighted to be working with Regen on this paper to provide insight into or thoghts in flexibility markets and to give some examples of or research into how a new distribted energy system cold operate. Merlin Hyman It is difficlt to exaggerate the speed and scale of the changes in or electricity system. Jst a few years ago power was generated in 50 or so large fossil fel and nclear power plants. There are now nearly one million renewable energy generators largely connected at the edge of the grid. Power reglarly flows p the distribtion network to the national transmission grid. Many hoseholders, landowners and bsinesses prodce their own power behind the meter, redcing their reliance on the electricity network. We are jst at the start of this energy transition. Technological innovation and the climate imperative are disrpting the market. Renewable energy is dropping in price and ot competing fossil fels. Technologies like electric vehicles and storage are developing rapidly whilst big data and ICT technologies enable innovative ways of matching spply and demand. Responding to these challenges will change every aspect of the way we generate, spply and se electricity. The age of exclsive control by big energy companies and central government is over; we mst maximise the ability of consmers to play an active role in managing their energy needs. Greg Clark, Secretary of State for Bsiness, Energy and Indstrial Strategy Working with the University of Bath the aim of this paper is to look at some of the key directions in which electricity generation, se and spply will change. We have chosen to examine innovation in three key areas: local energy markets enabling local balancing of spply and demand; flexible networks; and enabling technologies inclding ICT, storage and blockchains. 2 l Delivering Smart Energy Systems

3 Local Energy Markets Introdction In a highly distribted energy system matching spply and demand locally is an obvios step, redcing the need for expensive infrastrctre to carry electricity arond the contry, minimising losses, and enabling flexibility all of which cold redce cstomer bills. Local energy markets cold act as platforms where local balancing of spply and demand takes place. Trials of local energy markets are now beginning to emerge and the table below gives some examples of the different approaches. Research at the University of Bath is looking into the ideal design of the market that reflects the characteristics of local spply and flexibility, the key goals are to enable peer-to-peer trading of electricity at a local level and to enable mass cstomisation. Emerging models in the market Peer-to-peer trading Key featres We have all become familiar with online platforms like ebay that enable byers and sellers to connect directly. Electricity trading platforms follow the same principle of enabling peer-to-peer transactions. Generators can set the price for their power and consmers can identify the generators that they wold like to by from on an online platform. Generators can compete on price bt also other vales sch as providing renewable power and commnity ownership. In or crrent market set p an electricity spply licence is reqired for anyone selling power. Clarity on reglatory strctres will be key to enabling peer to peer trading and other approaches that Ofgem refers to as non-traditional bsiness models. Examples Selectricity is a trading platform (first tested throgh the Piclo trial) provided by Open Utility and Good Energy for bsiness cstomers that enables them to by electricity direct from their chosen renewable generators. Good Energy holds the electricity spply licence and ensres spply and demand are balanced. Vandebron in the Netherlands is another example of a trading platform. Vandebron claims that its consmers and prodcers both benefit by ctting ot the middleman. The consmers pay exactly what the generators price their power at. Vandebron then charges a flat sbscription fee to each side abot $12 a month to cover network and balancing costs. Delivering Smart Energy Systems l 3

4 Emerging models in the market Virtal networks Microgrids / private networks Key featres The se of smart meters to measre real-time se of electricity and/or generation opens p smarter ways of linking demand with spply. Virtal networks typically provide a commercial arrangement for balancing demand and spply at a local level. The commercial arrangement in a virtal network often takes the form of variable tariff rates to encorage consmption when local generation is generating. The greater the match between local generation and consmption the greater the potential for savings. A key area of debate is what generators and energy sers who are balancing their spply and demand at a local level shold pay towards the broader network. Private wire networks, or microgrids, enable a direct connection between generators and consmers. If the distribtor is providing less than 2.5 MW of power to domestic cstomers (roghly corresponding to 500 hoseholds), it can be classed as licence exempt spply. This makes it possible to ct ot the standard se of system charges and other obligations that apply to licensed sppliers. Private networks can either operate in parallel to or independent of the pblic network. Most remain connected to the pblic network to enable the export of excess generation or to top p when generation is low. Examples Energy Local is trialling a new model in Bethesda, Wales, that allows a local grop of domestic cstomers and a generator to form themselves into an Energy Local Clb (ELC). The power from local generators is then pooled and shared across the energy clb demand members. If members se the power when it is generated, they agree a price that the generator receives e.g. 7p/kWh. The Snshine Tariff trial in Wadebridge, Cornwall, incentivised cstomers to se electricity when a local solar farm was generating, sing a time of se tariff. The trial tested the concept that cheaper electricity at times of high generation cold relieve pressre on the local electricity network and enable additional generation to be connected. The Centrica Local Energy Market trial will explore the potential for cstomers to sell their flexible energy capacity to both the grid and the wholesale energy market throgh a virtal marketplace. Thameswey Energy is an nreglated pblic/ private joint ventre ESCo set p by Woking Borogh Concil in Thameswey provides electricity from a combined heat and power (CHP) generator to cstomers in Woking and Milton Keynes via a private wire network. The Centre for Alternative Technology microgrid connects demand to solar, wind, biomass and hydro generation onsite. When there is excess power, as mch as possible is stored in a battery and the rest is exported to the National Grid. Fishponds commnity land trst (CLT) is an example of one CLT that has installed onsite generation and a private network to directly spply tenants. 4 l Delivering Smart Energy Systems

5 University of Bath Peer-to-Peer trading Developing peer-to-peer local energy markets The changing energy system means that millions of bsinesses and homes, traditionally passive energy consmers, are becoming energy prosmers that can store, convert and/or generate energy, enabling them to be an active market actor. Each actor will make independent decisions to prse their own selfish or altristic goals. These emerging traits make conventional centralised control, dispatch and schedling tools no longer fit for prpose. To address these complex problems the University of Bath is leading a consortim addressing the strategic qestion: How to align the technical and market arrangements with diverse social reqirements, sch that selfish and independent goals prsed by millions of cstomers are aligned with the interests of the commnities and the system? The research is sing the principles of collaborative consmption/sharing economy to rethink how the energy market works and develop a prototype peer-to-peer (P2P) market platform. Distribted energy generation reqires a rethinking of how the energy market operates Delivering Smart Energy Systems l 5

6 Key research areas 1) Market architectre: Local P2P markets allow lots of fragmented byers & sellers to find and trade with each other, allowing individals/small bsinesses to compete with large incmbents. The approach of the research is to split the traditional market into a fixed central market and flexible local P2P markets that work in conjnction. The central market has controllable and predictable generators to set market price and delivers high qality spply to feed fixed demand. Local P2P markets match ncontrollable and npredictable generators with flexible demand at a local level, offering distribted approaches to address ncertainty and complexity. Price FLEXIBLE MARKET Spply Flexibility Price FIXED MARKET Spply Demand Qantity Qantity Horizontal and vertical energy spply provide more options to end cstomers. 6 l Delivering Smart Energy Systems

7 2) P2P pricing strategies: The key challenge to developing local P2P markets is to match large nmbers of spply and demand cstomers each with different priorities in terms of qantity, price and, critically, qality of power. Power generation varies in qality in terms of its availability on demand. The traditional market approach is not sited to matching varying spply qality with varying demand flexibility which leads to price spikes at times of peak demand. By enabling pricing strategies to recognise the qality as well as qantity of power at a local level a more efficient market can be developed. Price 100% Price 60% 40% 80% 100% 80% 60% 40% Qantity Qantity Qantity Qality Qantity Price Price By moving from one traditional market to mltiple markets, inclding P2P markets, the qality of spply can be priced to deliver more choice for flexible cstomers. SUPPLY CHOICES Spply-Demand Crve at 2:00 p.m. (Local market) DEMAND CHOICES Different spply options with differing availability need to be matched with differing cstomer demand flexibility. Local P2P energy markets can price different energy prodcts efficiently allowing large volme of energy to be traded at significantly cheaper cost. Delivering Smart Energy Systems l 7

8 The research has developed atomated and stochastic market pricing for renewable generation to make the local trading eqitable whilst sbstantially redcing the trading cost and time and offering lower prices for those whose demand is more flexible. Stochastic prices for a local PV generator for differing availability. 3) P2P market design: The crrent half-horly energy trading system is designed for traditional large-sale generation. Distribted renewable generation has very different characteristics. Identifying what prodct to sell over time and space is critical to prodct pricing and to matching spply with appropriate demand flexibility. A key area of research, therefore, is to examine in a local P2P market what prodcts are traded in terms of time of day, dration and size. For a typical solar energy prodcer, the energy prodct it can offer varies dramatically dring a half-horly period. Trading periods of five or even one minte cold be more sitable than half horly for distribted energy. Alternatively, another approach being considered wold be to have the trading period as event driven with the trading system atomatically determining appropriate prodct, price and energy volme based on instantaneos resorce characteristics. For demand flexibility, the degree of flexibility also varies very sbstantially from one prosmer to another dependent on technologies, behaviors and environment. Bath developed mathematical descriptions and eqations that can qantify and differentiate demand flexibility in terms of time, energy and continity so that the degree of flexibility can be nderstood and rewarded by markets. Matching spply and demand at a very local level has opportnities in accessing flexibility bt this needs to be balanced by recognising the inefficiencies introdced by transactions between varios levels of the system. The research is assessing whether energy trading platforms cold be operated for each LV distribtion sb-station feeder, (serving arond 200 homes in a typical rral location). The export/import from the platform shold be recognised by aggregators or sppliers who can either integrate the effects to the wider system or by expanding their services to inclde P2P matching. 8 l Delivering Smart Energy Systems

9 From the energy otpt for a PV generator over the year for typical energy prodcts can be identified from detailed analyses, characterised by time, dration, energy, and variability. Mathematical descriptions that can qantify demand flexibility in terms of time, energy and continity. Delivering Smart Energy Systems l 9

10 Pricing mst be established for energy prodcts and flexibility prodcts to widen participation to end cstomers. Load Flexibilty Matching Reslt Shiftable load Shiftable load Shiftable load Shiftable load Electric vehicle Electric vehicle Electric vehicle Energy storage Energy storage Energy storage Highly flexible cstomers are more likely to find a match that can take fll advantage of local intermittent generation. 10 l Delivering Smart Energy Systems

11 Flexible Networks Introdction To make maximm se of distribtion network assets and distribted energy resorces the way the distribtion network is operated needs to change. BEIS and Ofgem s plan for Upgrading or Energy System states that A key change is the contined evoltion of distribtion network operators (DNOs) to become more active in managing their networks as a system implementing innovative techniqes and exploring market-based soltions as alternatives to network reinforcement. This mode of operation wold see the DNOs operating as distribtion system operators (DSOs). The table below looks at some of the crrent changes of how the network is rn. The University of Bath research featred proposes a new approach, modelled on the telecoms indstry as to how capacity on the network cold be maximised. Emerging models in the market DNO to DSO Key featres The shift in the operation of the network is set ot in BEIS and Ofgems Upgrading or Energy System plan. The shift to DSO reqires DNOs to move on from simply operating a network to an active distribtion system that incldes demand, generation and other flexible distribted energy resorces. DSOs are likely to be expected to operate as a netral facilitator of an open and competitive market where third parties can se distribted energy resorces to provide services to the system. Examples All DNOs are working on the transition to a DSO. One trial of a mch more active approach is Western Power Distribtion s Flexible Power. Flexible Power helps bsinesses generate revene from their existing electrical flexibility. By adjsting electrical consmption or generation at key times companies can provide services that benefit them, the local network and the wider electricity system. The process has six stages: Consltation with the cstomer Adit cstomer energy se Recommend revene sorces Install eqipment Bid the cstomer in Send payments to the cstomer Delivering Smart Energy Systems l 11

12 Active Network Management Flexibility trading platforms One of the first featres of the shift to DSO has been the introdction by DNOs of Active Network Management (ANM). This approach is sed to atomatically control when electricity is exported, preventing the network being overloaded. When assessing reqests to connect generation to their network DNO look at the worst possible case of low demand and high generation in assessing whether reinforcement is needed to allow connections. ANM enables DNOs to allow connections and then to look at the actal flows on the network and disconnect generators when reqired. The ANM approach enables cstomers willing to have their distribted generation constrained when the network is overloaded to have their connections sped p and redce network reinforcement costs. A key challenge for maximising the potential of distribted energy resorces is providing a clear rote to market for anyone able to offer flexible generation, demand or storage. One approach set ot in BEIS call for evidence on A Smart and Flexible Energy System is for DSOs or independent local market platform operator(s) to collect bids and offers for flexibility actions from distribtion-connected providers in local areas (inclding from third parties sch as aggregators or commnity energy providers). These bids and offers wold then be sed at a local level to manage constraint and system reqirements. The local flexibility wold also need to be linked to the national system operation so that bids and offers from distribted providers cold be accepted by the local DSO and/or the SO to maximise their vale to the system. Active Network Management is being introdced by all DNOs, for example UK Power Networks annonced a new ANM area across parts of Cambridgeshire in Jne en/or-services/list-of-services/electricitygeneration/flexible-distribted-generation/ The Energy Networks Association has pblished a best practice gide to Active Network Management. energynetworks.org/news/pblications/ reports.html Centrica has commenced a three-year project trialling a Local Energy Market Platform in Cornwall. Western Power Distribtion is developing the software that will allow them to identify constrained areas of the network and by flexible energy demand or generation from the Local Energy Market platform. Centrica intend that they and other byers will be able to prchase energy and flexibility from homes and bsinesses. Local renewable generators sch as solar and wind farms will also be given the opportnity to sell energy via this platform l Delivering Smart Energy Systems

13 University of Bath Shared Network Access Developing new models for network access Distribted network operators (DNOs) revene has been largely determined by the amont of money spent on network investment each year. The development of distribted energy and behind the meter generation threaten a death spiral of the grid where a diminishing cstomer base leads to higher se of system charges reslting in frther loss of cstomers. University of Bath is carrying ot research into new bsiness models for DNOs to tackle the challenges of a declining cstomer base. Shared network access incentivises the incmbent DNOs to give p exclsive access to the network, leasing the spare or back p capacity to a licensed independent party. The ownership of assets will be retained by the incmbent DNO while competition will be introdced in the operation of the spare capacity. The independent parties who have license for (SNA) will act as secondary DNOs, tapping into the spare capacity in the network to provide flexible network services for flexible generation and demand. The principle is to take advantage of the largely idle spare capacity in the network to provide cheap alternatives for the integration of flexible demand. New operators with risk management expertise cold spply flexible demands with made-to-measre reliability contracts. This will more efficiently and effectively reflect and allocate to cstomers the real benefit of the ever-increasing electric demand flexibility. By leasing ot spare capacity DNO s cold enable new players with different risk profiles to maximise the se of the network Delivering Smart Energy Systems l 13

14 Key research areas 1. Leasing strategy and impact on network costs Rather than engaging in the complexity and risks of trying to develop models for sing spare capacity on the network, DNOs cold lease this capacity to a licenced third party that has a better nderstanding of the risk profile of cstomers connecting to the network. That third party will provide capacity to those who are willing to have their connection crtailed at certain times, if the price is right. This approach will develop a new income stream for DNOs and maximise the efficient se of the network. Modelling the shared network access approach indicates that a 10% network investment redction over conventional conterparts can be achieved for an 80% spply reliability. If 20% of the network is leased ot for 10% of annal network cost, then existing cstomers will benefit from a 10% redction in their network cost. Primary 20 MW 1000 typical cstomers (2kW/hose) Secondary 20 MW 1000 EVs (2kW/EV low flexibility) 2000 EVs (1kW/EV medim flexibility) 4000 EVs (0.5kW/EV high flexibility) Shared network access allows large volme of flexible demand connecting to the existing network at a fraction of traditional network cost. For those cstomers who can be very flexible in their generation or demand, network access cost serviced by the licensed third-party network provider wold be a fraction of the costs for traditional sers. This stems from both very low leasing cost and very high cstomer flexibility. For the same 20 MW leased network capacity, it can accommodate 4000 EVs if they have 4 hors charging flexibility and 2000 EVs if they have 2 hors flexibility. The cost of accommodating EVs with 4 hors flexibility wold be 10% of the traditional cstomer if the spare capacity can be leased at 40% of the annitised asset cost. 2. Forecasting A key capability needed to maximise the se of network capacity is accrate forecasting to identify spare capacity on all parts of the network. There is a lack of historical data on which to bild this bt by sing weather data, demographic information and big data analytical approaches the researchers are developing sophisticated forecasting tools. In particlarly, Bath developed new deep learning models to forecast the nexplainable ncertainties. The entire set of external variables are sed as inpts withot hman selection, able to nveil previos nknown interactions between internal and external variables and load demand. 14 l Delivering Smart Energy Systems

15 This example of load on a typical LV sbstation shows that there is no clear patterns for forecasting (Data from WPD). Third generation forecasting techniqes sing deep machine learning can identify and learn the key featres that are driving demand on the network. The reslts show that deep machine learning shows improvements in forecasting accracy compared with crrent models: a 39% improvement compared with ARIMA and a 28% improvement compared with SVM. Delivering Smart Energy Systems l 15

16 Enabling Technologies Introdction At its heart, the revoltion in or electricity system is driven by innovative technology. Distribted energy generation has developed at extraordinary speed. The ability to store energy to se when needed is becoming commercially viable. Information and commnication technologies are increasingly powerfl. The combination of these technologies offers the potential of an efficient distribted flexible energy system. The table below gives an overview of some of the key emerging technologies and where they are being deployed. To enable testing of different models the University of Bath has established a laboratory platform Enabling technologies Blockchain Key featres The energy system is increasingly evolving away from a handfl of centralised power stations to a portfolio of decentralised generators. Whilst the System Operator is responsible for ensing sfficient spply of power, and the health of the network, it is energy sppliers that manage the transactions between energy sers and generators. Many markets have benefitted from technology that connects spply to demand directly, removing the middle layer of brokerage and providing savings. The energy market is no different. Whilst a process already exists to connect generators to cstomers directly, it still involves a third party to broker the contract. Blockchain cold provide an alternative rote. Blockchain (the technology behind Bitcoin), allows peer to peer trading that is logged in a transactional database that everyone has a copy of and access to, meaning that no single party controls the data. Each new transaction is added to everyone s copy of the transaction list simltaneosly, making frad virtally impossible. For the energy sector, blockchain technology cold streamline one or more processes: Metering and billing Balancing and spply Emissions trading Smart contracting (system led spply and demand transactions) This technology cold drastically redce costs for generators and consmers. The challenge will be in realising the benefits of ctting the middle man whilst maintaining the investment needed to deliver and maintain a modern network. Examples TenneT Dtch-German tility TenneT and partners recently annonced two tility scale pilot projects a permissioned distribted energy blockchain and a residential solar/ storage and EV2grid integration This project will test the se of a blockchain database to access real-time views of battery storage/ PV generation/ev demand connected to the network. The blockchain essentially verifies the performance of each energy device, helping TenneT assess the transactional history of each participant and balance the system. 16 l Delivering Smart Energy Systems

17 Information & commnication technologies Big data is now common phraseology in many sectors, particlarly those that mst respond in real time. The ability to collect and analyse vast streams of data cold potentially transform the energy systems of the UK. As system operators acqire more power to balance energy flows sing storage or DSR, so the capacity of the network increases, allowing s to do more with less. Big data also means big investment. Crrently, mch of the UK electricity network has only limited data monitoring installed, let alone a system in place to analyse the data flows and initiate control. Smart meters represent one of the most significant investments in adopting a big data approach to the UK energy system. The roll ot of smart meters is still ongoing, bt cold be instrmental in enabling DSOs and the national Transmission System Operator to develop more accrate forecasting. Open LV EA technology has recently won 4.9m of fnding from Ofgem to trial a new data monitoring systems installed in low voltage sbstations. The aim is to stimlate new tools that will se the demand data generated by the system to meet their local energy needs. Deepmind National Grid is in talks with Google-owned Deepmind (a British based A.I developer) to apply machine learning to balancing energy flows. Blockcharge A joint ventre between RWE and Blockchain company Solck. it, Blockcharge combines a smart plg, smartphone app and a blockchain to allow cstomers to charge their electric car from any electrical socket. If they have their plg and smartphone, the phone app will negotiate the best contract and transact sing blockchain. This wold means drivers woldn t need charging stations for car charging any local otlet will do. Delivering Smart Energy Systems l 17

18 Emerging models in the market Aggregation Platforms Key featres In the context of smart energy, aggregation is an enabler for non-traditional or potentially smaller scale parties to be able to participate (and gain benefit from) incentive programmes or revene streams that they wold otherwise not be able to join. Aggregation allows parties who may have flexible energy demand or generation capacity (that can be called pon at will) to be added into a portfolio of other capacities, to be bid into balancing service contracts or capacity market contracts as a single entity. This type of aggregation can be achieved throgh partnering with third parties whose bsiness model is to aggregate capacity and ndertake the contract risk and operation themselves, known as aggregators. Alternatively, energy sppliers have started to see the vale in either promoting or actively recriting participants in DSR activity. Distribtion Network Operators (DNOs) are also actively looking at the potential to procre flexibility/balancing services in their respective licence areas. Some balancing isses (sch as voltage management and reactive power (VA R )) are mch more locational, compared to national balancing parameters (sch as network freqency (Hz)). Examples Aggregator company platforms/examples: LimeJmp Virtal Freqency Programme Kiwi Power Residential Demand Response app Upside Energy Virtal Energy StoreTM Flexitricity DSR aggregation REStore FlexTreoTM & FlexPondTM EnerNOC Energy intelligence / DSR Origami Energy Energy Technology Platform 18 l Delivering Smart Energy Systems

19 Storage There are a mltitde of opportnities for energy storage to provide both operational and financial benefit to the energy system. However, there are three roles of energy storage that are the core vales that energy storage can provide: Energy response charge/discharge qickly in response to grid isses Energy reserve store energy as a back-p or operating reserve Energy time shift / price shift charge now and store for se at a time when energy is most valable (operationally or de to price arbitrage) From assessing these core roles of storage, six key bsiness models for storage asset investment can be smmarised as follows: 1. Response service Providing higher vale ancillary services to transmission and distribtion network operators, inclding freqency response (FFR/EFR, ERPS) and voltage spport for network balancing. 2. Reserve service Specifically to provide short/medim term reserve capacity for network balancing, sch as Short Term Operating Reserve (STOR), Fast Reserve or the Capacity Market. 3. Commercial and indstrial high energy prosmers Located with a high energy ser (with or withot on site generation) to avoid peak energy costs and peak transmission/ distribtion network charges, while providing energy continity. 4. Domestic and commnity own-se Domestic, commnity or small commercial scale storage, designed to maximise the self-se of generated electricity to avoid peak electricity costs. i.e. rooftop PV and storage in the home. 5. Generation co-location Storage co-located with variable energy generation in order to a) price/time shift or b) peak shave so as to avoid grid crtailment. UKPN Smarter Network Storage (Li-ion) WPD/RES/BSR Copley Wood WPD (Li-ion) SSEPD Orkney Energy Storage Park (Li-ion) Anesco Slepe Farm (Li-ion) AES - Kilroot (Li-ion) RedT - Gigha (Flow Battery) Electric Montain - Dinorwig (Pmped Hydro) Viridor - Highview Power Storage (LAES) Delivering Smart Energy Systems l 19

20 University of Bath Local DC Research and development platform The University of Bath has developed a laboratory platform to stdy and research the interplays between consmers, technologies and the grid nder differing commercial arrangements and innovative bsiness models. The platform is based on SoLa Bristol Low Carbon Network Fnd project led by Western Power Distribtion that trialled new methods to connect solar PV with in-home batteries sing a local DC network, variable tariffs and energy management systems. The key goal is to encorage demand redction and demand shifting at cstomers premises to relieve network congestion whilst taking advantage of cheaper grid energy to redcing cstomers energy bills. The platform is crrently being sed to develop and trial different commercial models, energy integration, energy management systems and load forecasting techniqes for batteries to enable local demand responses that will spport the grid and deliver vale to renewable generation. Key research areas Commercial battery models The research has identified the commercial revene streams and is developing ftre market streams that batteries cold access and then portion battery capacity into mltiple shares to target different revene streams. Battery revene opportnities over different timescales. 20 l Delivering Smart Energy Systems

21 Efficient energy management system A central challenge of maximising the vale of storage is to design an efficient battery energy management system that can change its charging/discharging strategy in response to a variety of different conditions, sch as weather, market, grid and local conditions. The aim is to maximise the retrn for the cstomers assets whilst ensring the assets safety and offer spport to the whole spply systems across energy and network. The research has developed a highly efficient optimal battery charging/discharging strategy that can arbitrage between local spply and central spply, between efficiency losses and financial gain by shifting energy over time and between ncertainty and flexibility. The battery charging/discharging algorithm developed is highly efficient at maximising the vale of the battery to the LV network and in taking advantage of cheaper grid energy. Maximising vale to the LV network A key challenge is the ncertainty of demand on the distribtion network particlarly in the last mile. Understanding the minimm demand response reqired to show meaningfl network impact is, therefore, challenging. A key qestion for the research has been how to nderstand what cstomer response cold reliably meet the operational targets of the distribtion system. The University of Bath has developed a decentralised energy management system that can achieve global impact from energy management at the individal hosehold level. Level of ncertainty for two typical LV sbstations. By sing a probabilistic techniqe the research has identified the minimm reqired DSR cstomer engagement to spport operational targets of distribtion system. Delivering Smart Energy Systems l 21

22 22 l Delivering Smart Energy Systems Notes:

23 Delivering Smart Energy Systems l 23 Notes:

24