Chapter 21
Energy Storage Integration
Philip Taylor*
Charalampos Patsios**
Stalin Munoz Vaca**
David Greenwood**
Abstract
As fundamental storage technologies improve in terms of performance and cost their use on electricity networks will become increasingly attractive. It is likely that the business cases that emerge will continue to be sensitive to the ability to capture value streams from delivering more than one service. Successful service delivery will be dependent on the availability of appropriate and effective planning and operating schemes. Validation of these schemes through demonstration projects is an essential step in building the experience and knowledge required to widely deploy energy storage in electricity networks. The chapter covers a range of the work that has been undertaken by researchers, working with the electricity industry, to advance the understanding of energy storage. The importance of policy and markets in creating an acceptable commercial environment is described. Methods used for network planning of energy storage installations are outlined. Operational strategies, both for individual storage units, multiple units, and in combination with other technologies are provided. Examples of demonstration projects and their outcomes illustrate postal progress in deploying energy storage live networks. Finally, advanced methods for integrated storage modeling are explained, with examples of the findings that can be made. Under current regulatory and electricity market conditions, conventional solutions, such as fossil fuel–based peaking plants for covering peak electricity demand, are seen as cheaper technologies when compared with energy storage systems (ESS). However, this does not necessarily represent the true cost and value of ESS. This circumstance is expected to change as ESS technologies continue to advance, capital costs reduce, deployment experience increases, and there is the opportunity to reconcile multiple value streams. To enable grid-connected energy storage to flourish, effort is required in a number of areas: (1) Energy policy decisions must be reviewed and ESS policies must be aligned with those for renewable energy systems (RES) so that the ability of ESS to add RES capacity to the system is rewarded. (2) Regulatory rules need to be examined; a new asset class and associated regulation specifically for ESS is advisable; and standardized evaluation methods for determining the value of storage in power systems would reduce investor uncertainty. (3) Updated ancillary services markets to provide adequate compensation for technologies that can respond quickly and with high accuracy. (4) Operation schemes should be investigated; the capture of multiple value streams is dependent on the ability to control storage in a way that can balance competing requirements appropriately. (5) Roadmap for ESS deployment on the grid; if ESS is considered as a potential solution, it is important that plans, targets, and goals for the use of ESS are established. (6) Development of sophisticated modeling tools; the properties of storage media and demands placed by grid control requirements are complex and at times contradictory. Integrated modeling of these will allow the development of schemes that are able to be more cost effective. (7) Deployment of projects; an increasing number of storage installations are in operation, but the multifaceted nature of storage means that further novel applications still need to be deployed. Addressing these seven areas will mean that the potential of energy storage for supporting electricity networks can be practically realized.
Keywords
distribution networks
energy storage demonstration
planning
operation
integrated modeling
1. Introduction
The chapter seeks to cover the essential aspects of the network integration of electrical energy storage (EES) systems. The chapter covers energy storage policy and markets, energy storage planning and operation, demonstration projects involving network integration of energy storage and energy storage modeling. The chapter finishes by drawing conclusions about the current state of energy storage deployment and future requirements for research, development, and deployment. Interest in electrical ESS is increasing as the opportunities for their application become more compelling in an electricity industry with a backdrop of aging assets, increasing distributed generation, and a desire to transform networks into smart grids.
Energy storage is not in itself new; electrical ESS have been in use since at least 1870 when Victorian industrialist, Lord Armstrong, built one of the world’s first hydroelectric power stations at Cragside in Northumberland (United Kingdom). In hydroelectric schemes, the penstock valve regulates the conversion of potential energy held by water in an upper reservoir into electrical energy by a turbine-generator set. The storage capacity of a scheme is determined by the volume of water available in the reservoir and the power output by the rating of the generator.
It was identified as early as 1959 that to make best use of renewable energy resources with a meteorologically dependent output a storage element to the overall system would increase the energy yield. As well as increasing yield the ability to add dependability to renewable resources has been widely investigated. Despite this long track record a number of research challenges remain to be overcome if energy storage is to become ubiquitous and live up to its potential.
The current global implementation of energy storage in power systems is relatively small but continuously growing with approximately 665 deployed projects recorded as of 2012 [1]. Worldwide grid energy storage capacity was estimated at 152 GW (including projects announced, funded, under construction, and deployed), of which 99% are attributed to pumped hydro schemes and the remaining installations are new nontraditional storage systems (such as batteries and flywheels). ESS are designed to meet performance criteria spanning timescales from milliseconds to days and even seasons and from watt-hours (W h) to gigawatt-hours (GW h).
A number of commercial and regulatory barriers exist to the large-scale takeup of electrical ESS in power systems, including [2]:
• the high implementation costs for energy storage in many cases make it uneconomical compared with conventional solutions;
• the unbundled electricity system (prevalent in many countries) which lacks transparency means stakeholders cannot determine the full value of energy storage;
• the undetermined asset class for energy storage, which functions as both generation and demand, means potential investors cannot always realize the benefits of storage; and
• low electricity market liquidity, changing market conditions, and a lack of common standards and procedures.
Electricity distribution networks have entered a period of considerable change, driven by several interconnected factors: aging network assets, installation of distributed generators, carbon reduction targets, regulatory incentives, and the availability of new technologies [3,4]. In this climate the use of distributed storage has reemerged as an area of considerable interest. The end of this period of transition will be signaled by the successful establishment of the technology and practices that must go together to create what is termed the Smart Grid. The UK Electricity Networks Strategy Group (ENSG) provide a useful definition of the term [5]: a smart grid as part of an electricity power system can intelligently integrate the actions of all users connected to it—generators, consumers, and those that do both—in order to deliver efficiently sustainable, economic, and secure electricity supplies.
It is clear that the integration of electrical ESS into electrical networks is a key enabler for smart grids and decarbonization of the electricity industry. The chapter describes the key issues which must be considered and addressed when attempting to integrate energy storage into electrical networks.
2. Energy policy and markets
Governments, utilities, regulators, and other electricity stakeholders are all interested in the role of ESS in providing solutions in evolving and future power systems due to their versatility in providing power and energy capacity. As policies, electricity markets, and regulatory frameworks are constantly evolving, so is ESS, which although in its infancy will mature in the years ahead. It is estimated that the global demand for ESS will be £72 billion by 2017 [1], and in the United Kingdom, for example, bulk ESS has been projected to provide annual benefits of £120 million by 2020, £2 billion by 2030, and over £10 billion by 2050 to integrate low-carbon technologies (LCTs) to the grid (with similar achievable benefits for distributed ESS) [6]. The investment potential in the United Kingdom can be applicable to power sectors in numerous countries facing similar issues. Nevertheless, the unconventional operation and different functions of ESS complicates their operation under the current regulatory and market structures.
2.1. Background
2.1.1. Regulation
To increase competitiveness, provide higher quality services to consumers, and drive down costs in the power sector, deregulation was introduced for generation and supply functions [7]. In a restructured and deregulated electricity system, generation and supply functions are generally classed as competitive while the transmission and distribution (T&D) networks are regarded as natural monopolies and are regulated [8]. Regulation is used as a tool to drive down the cost of electricity and ensure a low electricity tariff for customers, provide a return on investment for electricity network stakeholders involved with T&D, and provide incentives to T&D companies to improve both network and operating efficiencies to the benefit of customers.
2.1.2. Electricity Market
Wholesale electricity markets usually operate as a centralized market (power pool) or decentralized market (bilateral contracts) [9]. The markets in a liberalized electricity system are futures, spot (day-ahead and intraday), balancing, ancillary services, and retail. In the wholesale forward market, short-term contracts are carried out in the spot market (day-ahead and intraday markets) and long-term contracts are made in the futures market, which covers trades for a week up to a year. To maintain grid frequency and system stability, supply and demand has to be balanced constantly in real time due to the lack of storage capacity in power systems. System balancing is carried out via the balancing and ancillary services market to account for shortfalls in the spot market.
2.2. Business Models for Using Energy Storage
A major issue affecting the wider implementation of ESS is the higher costs they add to the already expensive T&D networks or renewable energy systems (RES) deployments, which often renders them uneconomical if used for a single application when compared against alternative conventional solutions. Thus, developing a viable business model for the provision of multiple functions is important for the success of ESS.
In an unbundled power sector, ESS could be used (if regulation permits) for competitive (deregulated) services in the wholesale energy market (day-ahead and intraday), balancing and ancillary services markets, and capacity markets to maximize value across the electricity value chain. According to Pomper [10], ESS owners and providers can be categorized into six types: merchant providers, transmission system operators, distribution system operators, customer group, or contract storage providers. The ownership types, regulatory frameworks, and location of the ESS all influence the business model, which can either be regulated and/or competitive.
ESS used under the regulated business model would provide a guaranteed revenue source as this would be fixed based on contractual terms for services provided or if owned by a regulated network operator would lead to guaranteed cost recovery. Conversely, the deregulated or competitive business model can be used to participate competitively in the electricity markets and can additionally be used to provide regulated services without major interference [10,11].
2.3. Review of National Policies, Regulation, and Electricity Market Arrangements Supporting Storage
Europe. ESS investment in Europe covers over 20% of the ESS market worldwide [1,12,13]. The European Commission (EC) developed the Strategy Energy Technology Plan (SET-Plan) for developing and implementing an EU energy technology policy for the transition to a low-carbon economy [14]. The aim of the SET-Plan is to change the EC’s approach toward investing in research, development, and demonstration (RD&D) activities for a low-carbon economy and it includes materials for ESS pathways for energy storage in the United Kingdom [15].
United Kingdom. The UK government has a 15% renewables target by 2020 and plans for the increased electrification of transportation and heating by 2030. Thus, the government has identified ESS, interconnection, and demand response as crucial in enabling the United Kingdom to reach its targets for transforming the electricity system by the year 2050 [16]. However, there is no clarity on the future role of ESS in the United Kingdom and consequently no specific regulation for ESS. There are no specific license conditions for ownership and operation of ESS, which functions as a load or generator. At present, ESS is considered as a generator under license conditions [17].
Scandinavia. ESS has been identified as a pivotal element in the Danish 2050 energy vision [18]. The present regulation in Denmark treats ESS as load, hence ESS is liable to grid charges for load. In Norway, there are grid charges for pumped hydro storage as load or generator with an additional charge for energy consumption during peak periods [17]. At present, there is no specific regulation or electricity market change for ESS in Norway and Denmark but there are future plans for using storage.
Germany. Germany’s electricity market is Europe’s largest [19]. ESS is considered to be a key component in the country’s move toward a reliable, economically stable, and efficient power system. A report on European regulatory aspects for electricity storage [20] concluded that the lack of regulations, opportunities, and mechanisms to support the competitive use of ESS is affecting the uptake of ESS in Germany.
Spain and Italy. There is no specific regulation for ESS in Spain and legislative initiatives have been restricted to the Canary Islands where compensation is realized through regulated capacity and energy payments. In Italy the rapid increase in RES led to new legislative initiatives and proposals to be passed. The Legislative Decree 28/11 implementing directive 2009/28/EC calls on Terna, the Italian transmission system operator (TSO), to identify network reinforcements, including ESS, to enable energy from RES to be fully dispatched [21].
Australia. Australia has high carbon emission reduction targets as the country has the highest per capita GHG emissions in the Organization for Economic Co-operation and Development (OECD) and one of the highest globally [22]. There is currently a target of 20% electricity production from RES by 2020 (as illustrated in Fig. 21.1), which is expected to help reduce GHG emissions by 5% [23–26]. Nonetheless, the increase in RES has not yet brought about an interest in ESS, which currently do not participate in the energy market. Due to the lack of support, experience, and uncertainty of its future role, utilities do not include ESS in network plans and are unsure on how to recover costs for investing in ESS.
Figure 21.1 Renewable energy sector (RES-E) 2020 percentage targets for energy consumption in Australia, China, and Japan. Source: [21].
China. The Chinese electricity sector is vertically integrated with state-owned monopolies but there are currently plans to reform the power sector by unbundling transmission and distribution and deregulating the electricity market [27]. The market for ESS use is motivated by the need to increase the efficiency of the grid by the integration of RES. As the Chinese electricity market is not competitive, policies would be the main drivers for developing ESS. However, the current lack of national policies supporting ESS is a major barrier to national uptake of the technology [28,29].
Japan. The power generation and retail sectors of the power industry are liberalized and controlled by 10 vertically integrated power companies [30]. After the Fukushima nuclear disaster the government’s interest in the use of ESS in providing security of supply has increased. There is a short- to medium-term target for 15% ESS capacity to be deployed on the grid [31].
United States. The investment in ESS is growing and was encouraged by the Energy Independence and Security Act of 2007. This identified the use of advanced electricity storage and peak-shaving technologies as a means of modernizing the grid to maintain reliable and secure electricity infrastructure and to meet growth in demand [13,32]. Since ESS do not fall under the conventional functions of generation, transmission, or distribution, the Federal Energy Regulatory Commission (FERC) individually addresses issues with the classification of ESS for use on the grid [30]. A major challenge for FERC is developing and adapting markets in deregulated states and creating proper evaluation frameworks in regulated states to allow ESS technologies to have economic value for the range of benefits that they can provide [33,34].
2.4. Regulation, Electricity Markets, and Their Impact on Storage Implementation
The regulatory barriers restricting the uptake of ESS are to a large extent dependent on the extent of unbundling practised. In countries where unbundling has not been fully realized and there is vertical integration, it is easier for utilities to deploy ESS across the power system to support the grid and to also use ESS commercially in the electricity market. On the other hand, in an unbundled system, the benefits derived from implementing ESS are more challenging to determine and accomplish because there are multiple actors involved from generation to consumers with different goals, practices, and regulation systems in place.
2.4.1. Storage Regulatory Barriers
Renewable integration policies. ESS could be used to store excess energy from RES and substitute for grid expansion. However, there is little incentive for investment in ESS because of the high priority and financial compensation provided to renewable generators to curtail excess energy.
Transmission and distribution use of system charges. Regulation determines whether T&D use of system charges should apply to ESS when used to provide services on the grid. Presently, ESS used on the grid is subject to T&D charges as a generator, consumer, or both, depending on the country.
Undetermined asset classification. ESS are multifunctional and can serve as a generator, transmission, or distribution asset, or as an end-user, depending on the required end-goal. Consequently, ESS asset classification is undetermined under present regulatory conditions and this directly affects eligibility for ESS asset ownership, grid tariffs, and cost recovery for regulated assets.
Lack of framework and incentives for storage service provision to network operators. There are no incentives or rewards in place for improved power quality, and power quality benefits are difficult to quantify [35]. The benefits ESS provide by improving capacity utilization of the grid and increasing the efficiency of centralized generators are also difficult to quantify.
Unwillingness to take risks or innovate. Although some ESS technologies are established, most technologies (other than Pumped Hydro Storage (PHS)) are still developing for use on the grid. Hence the lack of experience and high investment costs make it a risky venture.
Lack of standards and practices. Most ESS with the exception of PHS are relatively new and developing technologies (e.g., compressed air energy storage, hydrogen storage) with minimal deployments. This has resulted in a lack of necessary standards and practices to carry out thorough economic assessment, system design, and deployment.
Policies for other competing technologies or solutions. Current policies favor established technologies (e.g., interconnections, gas peaking power plants) over the use of ESS which have limited operational experience holding back the growth of ESS implementations.
Investment dilemma. The difficulty in determining the wide range of benefits across the grid makes it difficult to quantify the overall value of an ESS investment. This affects the profitability of investing in ESS and is especially the case for independent ESS owners.
Energy storage not being considered as part of RES schemes. The production of electricity from ESS connected to the grid may or may not be from RES. This creates a difficulty in trying to include the benefit of ESS under RES subsidies.
No benefit for controlled and dispatchable RES. Generally, generation-based support mechanisms (market premiums or feed-in tariffs) and priority dispatch are part of regulatory frameworks and policies to increase the uptake of RES. However, these mechanisms do not include and compensate for the controlled dispatch of renewable energy to meet demand and supply variations on the grid [20].
2.4.2. Storage Market Design Barriers
Limitations on service participation. ESS owners may be unclear of the future state of charge (SOC) of ESS when they are participating in the balancing or reserves market. This may prevent ESS owners from being involved in the spot market and in providing grid support services because it would be difficult to guarantee use in the balancing or reserves market if ESS are used for several services.
Lack of market liquidity. A liberalized electricity market that promotes competition favors a liquid wholesale market. However, bigger generators engage in bilateral contracts to mitigate issues that arise as a result of volatile prices in the spot market, which is currently affecting Distributed Generation (DG) operators in the EU [36]. This leads to low electricity market liquidity which is an entry barrier for ESS owners because it limits access and results in an unreliable market [37].
Market operation requirements and market fees. Satisfying the requirements of the spot market could be difficult if ESS are being used in multiple energy markets. Confirmation is required in close to real time for the provision of balancing and other ancillary services, which conflicts with wholesale market requirements, in which participants confirm their position ahead of real time in futures, day-ahead, or hour-ahead markets.
Decline in spread of peak and off-peak energy prices. The spread of energy prices during peak and off-peak periods provides an avenue for ESS owners to gain revenues from energy arbitrage. The price spread is affected by factors including the demand and generation mix and unpredictable fuel and carbon dioxide prices [38,39].
Unfair advantage provided to regulated utilities. The use of ESS by natural monopolies could complicate electricity market operation as it can provide regulated network operators with a way to influence the electricity market price and provide a biased advantage, which goes against the principles of unbundling.
Market price control mechanisms. Price control mechanisms enacted in different countries may affect the revenues ESS can make from arbitrage. As ESS may often operate for shorter periods during the year, compared with conventional generation, the opportunity to recover investment costs during periods of volatility in the markets is important. A price cap will create uncertainties that will significantly affect the business case for investing in ESS.
Wholesale and retail price market distortion. T&D operators have the potential to distort the electricity market by participating in the wholesale and retail markets while also obtaining regulated revenue on ESS, as a network asset, placing them at an advantage against other ESS or generation owners.
Undifferentiated remuneration for reserve and other ancillary services. ESS are compensated in the same way as traditional regulation service providers despite the additional benefits that their accuracy, high responsiveness, and rapid ramp rate can provide.
Penalties for not meeting scheduled energy dispatches. Using storage under a business model where it provides regulated and competitive services would be difficult to control. This is crucial as market rules place financial penalties on operators if an ESS is contracted to provide reserve services or electricity in the wholesale market but does not have enough available energy due to it being used for other services.
Value assessment from market operations. The method of assessing the potential revenues from ESS providing services in different electricity markets is complex because of the associated risks and uncertainties of changing market conditions.
Size requirements for ancillary services markets. There are limitations placed on the minimum duration and size of generation that can participate in providing regulation services. There are also limitations on the energy capacity in reserve markets. This limits participation from operators with smaller sized ESS.
3. Energy storage planning
Planning the use of energy storage in electrical networks is an important task which involves offline analysis to determine the optimal rating, capacity, location, voltage level, and service provision for ESS. Network operators are interested in the costs and benefits of different technologies to manage their assets.
Reverse power flow and voltage rise are related events which are most extreme when generation is highest and demand is lowest [40]. These both become more problematic as the penetration of rooftop PV increases and can limit the amount of distributed generation that should be installed in a network area [41]. Consumers connected to the LV network may have no direct problems as a consequence of reverse power flow, but this can affect voltage regulation on the medium voltage network including additional cycling of tap-changing transformers [42]. Reverse power flow is currently evident in the UK distribution network, as metered data from an LV transformer shows (Fig. 21.1). Voltage must not exceed 10% above or 6% below 230 V under UK regulation [43].
It is estimated £32 billion of investment is needed to mitigate the effects of distributed generation in the UK electrical network. To manage the network without directly interfering with generation or customer demand, network operators can either reduce network impedance (reconductor), add discretionary loads, demand-side management, or energy storage [42]. Indeed, under the new price control scheme (RIIO: revenue, incentives, innovation, and outputs) there is a financial incentive for distribution network operators (DNOs) to invest in new technologies and techniques such as energy storage [44]. In a competitive industry, there is a need to assess the cost implications of these innovative technologies relative to traditional mitigation methods. Energy storage is widely considered to be a technically viable solution to the problems expected in the distribution network, eg, in [45]; however there are few industrial distributed storage projects, costs are high, and DNOs do not necessarily have the experience to plan for new technologies.
EES technologies can generally be split into two broad categories. Utility or bulk-scale energy storage, such as pumped hydro and compressed air, are capable of delivering several megawatts of power over (1–8) h and due to cost and geological restrictions are suitable for transmission applications [46]. Distributed storage systems typically deliver smaller amounts of power for a similar period to utility storage but can be scaled in terms of rating, location, and capacity [47].
As summarized in [45], storage can offer a number of benefits to DNOs including voltage support, power flow management, restoration, network management, and compliance with regulatory requirements. Reverse power flow and voltage rise can be managed using storage to absorb power from the generators. Peak shaving (PS) requires discharge of stored energy into the network to reduce the loading on transformers and cables. To reduce reverse power flow or to peak-shave requires a specific amount of energy to be supplied or absorbed. Assuming thermal limits are not exceeded, power/energy specifications of the storage are unaffected by its location in the network.
However, the power required to manage a voltage problem is directly affected by the location of the storage unit [48,49]. As such, to reduce the capacity and power required to solve a voltage problem, storage should be distributed at many nodes within the network. It is often suggested that locating the storage within the network (in properties or on the street) will allow the greatest ability to reduce the power needed to solve the voltage problem. In real networks, however, there may be hundreds of nodes (busbars, customer connection points) where an energy storage unit could be connected. If multiple storage units are proposed the number of feasible combinations of energy storage increases rapidly; for example, there are 2.25 × 1016 ways of locating 10 storage units among 200 feasible locations.
3.1. Heuristic Techniques
Due to the complexity of this problem heuristic approaches are often considered for locating distributed storage. A number of papers consider heuristic approaches in the location, sizing, or operation of energy storage. In [50] a Tabu search approach is used for sizing energy storage by considering unit commitment. In [51] the authors use a genetic algorithm to locate superconducting magnetic energy storage to maximize the voltage stability index. In [19] three cost-based heuristics are shown for managing voltage rise in LV networks and the authors find that deterministic approaches are not as good as stochastic methods because they are unable to search the entire problem space. In [52] a genetic algorithm is used to locate and size a single energy storage unit to achieve benefits in reducing loss, voltage deviation, and costs. In [53] a genetic algorithm is combined with a sequential quadratic programming approach to locate capacitors and energy storage in a medium-voltage (MV) smart grid. In [47] a multiobjective algorithm is used to locate and size storage units in a 34 bus, 24 kV network. Objectives include reduction of storage power and capacity, minimizing the probability of voltage deviations, maximization of arbitrage revenue, and minimization of lost ancillary service opportunities. The heuristic used in [47] builds on work in [54] where wind, PV, and CHP units are located in a distribution network using a genetic algorithm.
Although global and local search methods have been applied to distribution networks in the literature, further consideration is needed into how their application can provide relevant results to DNOs in relation to distributed energy storage. This particularly applies in the area of planning given uncertainty and a lack of control of the location of distributed generation.
3.2. Probabilistic Techniques
When a primary substation reaches its capacity limit the standard solution is to reinforce the network with additional circuit capacity. Under the right conditions the required additional peak capacity can be provided from EES [55], real-time thermal ratings (RTTR) [56], or a combination of both. Probabilistic methods for calculating the size of an EES system for a demand peak-shaving application are described in the following text. The impact of both power and energy capacity are considered, along with the reliability of the energy storage.
3.2.1. Peak Shaving
Peak shaving (PS) refers to the reduction of electricity demand at times of peak consumption. Electricity demand varies throughout the day; in the United Kingdom this peak typically occurs in the early evening. In the majority of cases, peak demand only occurs for a small fraction of the time [57], but the generation, transmission, and distribution systems are constrained by it. A PS scheme attempts to reduce the demand peak either literally, through demand-side response [58] or offsetting the demand by supplying power locally through distributed generation or energy storage [59]. It is necessary to determine how large an EES system, in terms of both power rating and energy capacity, should be used to offset peak demand. EES has several potential advantages over conventional reinforcement; it can help solve other problems, for example, overvoltage on the local network, it does not require the same long planning consultation, nor does it result in the same long-term lock-in; it can participate in balancing, reserve, and frequency service markets.
Fig. 21.2 shows the demand over 24 h at a primary substation (33 kV/11 kV) in the United Kingdom. Between 1700 and 2000 the demand exceeds the overhead line limit, so the EES would need to provide PS for those 3 h. The peak exceeds the demand by around 4 MV A, so the EES would need a converter rated at 4 MV A to successfully offset the peak. The total area between the demand curve and the line rating is around 8 MW h; this is the energy capacity that would be required for the EES to meet the PS demand.
Figure 21.2 Peak shaving needs to take place when the existing infrastructure cannot support the demand peak.
A storage system would need to meet both power and energy requirements to successfully offset the peak.
A storage system would need to meet both power and energy requirements to successfully offset the peak.
Electricity demand is highly variable, depending on time of day, day of week, month and season, as well as physical variables such as temperature and whether it is light or dark. Consequently, PS time, power, and energy will all vary from day to day. The ability to predict this demand is a crucial aspect of using EES for PS because sufficient energy will need to be stored in the EES prior to the PS event.
For this example system, increasing the power rating of the EES only provided benefits up to around 4 MW, while energy capacity continued to provide improvements up to 20 MW h. The addition of RTTR proved more beneficial in a load growth scenario because demand is likely to be above the static rating for a greater proportion of the time. RTTR also reduces utilization of the EES for PS, either freeing it to perform more commercial services or extending its operational lifetime. This extension is particularly useful given that the combination of RTTR and EES is likely to solve a network problem for considerably longer than either would in isolation.
3.3. Planning Storage for Security of Supply
EES can bring benefits on occasions when the network is disrupted. Under these circumstances the number of customers who can be resupplied at any time, and especially at times of peak load, is likely to be limited by the thermal ratings and consequent maximum capacity of certain critical sections of the network. Increasing this number would constitute an improvement in quality of service and may be required by the national industry regulator. In some circumstances, this requirement could involve costly capital expenditure, for example, to reconductor an overhead line or underground cable to a higher static thermal rating.
However, if supplementary energy supplies are made available from EES, then the number of customers supplied during such an event can also be increased. This can be achieved by smart deployment of a battery storage system.
This methodology is illustrated by a case study, described in the following section, which also calculates the supply shortfall that would be expected under faulted conditions. The potential impact of EES is then outlined, and the potential of different-size EES systems to reduce the risk of customer disconnection is calculated. The final section draws conclusions from the case study, in particular as regards the potentially wider use of EES technology and the associated methodology.
3.3.1. Case Study
Primary substations “A” and “B ” together serve over 16 000 customers in the north of England, with a present peak demand of over 34 MW. They are supplied by two independent teed 33 kV circuits, as shown in Fig. 21.3. These supply circuits each consist of underground cable for the first 2.9 km, followed by 1.6 km of overhead line to the tee. These sections of overhead line are the most critical as regards static ratings.
Figure 21.3 Schematic of supply circuits.
In the event of a fault on one of the two circuits the remaining circuit would be required to carry the full load to both primaries. The critical section of both circuits is 175 mm2 aluminum conductor steel reinforced (ACSR) overhead line, with static ratings of 30.8 MV A (winter), 28.6 MV A (spring/autumn), and 24.7 MV A (summer). Analysis of actual half-hourly load data for the 12 months from Aug. 2011 to Jul. 2012 indicates that the summation of load at both primaries reached peak values of 33.28 MV A in winter (Jan. 18), 29.29 MV A in spring (Apr. 25), and 27.86 MV A in summer (May 1), all of which are in excess of the single-circuit static rating.
3.3.2. Potential Impact of Electrical Energy Storage
The following analysis assumes the connection of a 5.0 MW h, 2.5 MV A EES system to the 11 kV busbars at Primary “B.” The ability of the EES system alone to secure the shortfall is a function of both the power rating of the converters and the energy storage capacity of the unit. For example, on Jan. 18, 2019, the expected worst day of winter, the shortfall would last from 1700 until 2030 and would be in excess of 2.5 MV A from (1730–1900) inclusive, a total of 2 h, as shown in Fig. 21.4. With peak converter power of 2.5 MV A the shortfall could be fully secured for 0.5 h, then partly secured for 2.0 h, then fully secured for a further 1.5 h.
Figure 21.4 Shortfall predicted, Jan. 18, 2019.
The total energy shortfall on Jan. 18 is 8.75 MW h, which is almost double the assumed battery capacity of 5.0 MW h. As regards energy, only the first 1.8 h of shortfall could be secured.
For smaller power shortfalls the converter can be operated at below capacity and thus support an energy shortfall of longer duration.
3.3.3. Effect of Electrical Energy Storage System Size
Analysis of power and energy shortfalls on the 146 days per year on which an energy shortfall could be expected for part of the day in the event of (n – 1) loss of a single circuit has been carried out, and the results are shown in Table 21.1. The duration of the shortfall period, which is not shown in Table 21.1, ranges over those 146 days from a minimum of 0.5 h up to a maximum of 13 h. It is assumed that the time between shortfalls (at least 11 h with load levels well below peak) allows the EES to be fully recharged for any combination of converter size and storage capacity.
Table 21.1
Number of Days with Specified Energy and Peak Power Shortfalls
| Peak power shortfall/MV A | |||||||
| Energy shortfall/MW h | 0.0–0.5 | 0.5–1.0 | 1.0–1.5 | 1.5–2.0 | 2.0–2.5 | 2.5–3.0 | Over 3.0 |
| 0.0–1.0 | 22 | 18 | 3 | ||||
| 1.0–2.0 | 1 | 5 | 21 | ||||
| 2.0–3.0 | 3 | 5 | 2 | ||||
| 3.0–4.0 | 2 | 4 | 7 | ||||
| 4.0–5.0 | 2 | 1 | 7 | 1 | |||
| 5.0–6.0 | 2 | 1 | 5 | 1 | |||
| 6.0–7.0 | 1 | 3 | 2 | 3 | 1 | ||
| 7.0–8.0 | 1 | 2 | 3 | ||||
| 8.0–9.0 | 2 | 1 | 2 | ||||
| 9.0–10.0 | 1 | 1 | |||||
| Over 10.0 | 1 | 6 | 3 | ||||
The numbers in the body of Table 21.1 indicate the number of days on which energy and peak power shortfalls fall within a given range. So, for example, there were 21 days within the projected year 2018–19 when, in the event of an (n – 1) fault, the energy shortfall would be between (1.0–2.0) MW h and the peak power shortfall would be between (1.0–1.5) MV A.
Table 21.1 can be used to evaluate the effectiveness of a converter and battery of the size installed at the trial site (2.5 MV A and 5.0 MW h). By totaling cells, it can be seen that on 103 of the 146 days this EES system would be able to meet the whole shortfall. On 19 of the remaining days its converter power rating would be sufficient, but not its energy storage capacity. On one of the remaining days the energy storage capacity would be sufficient, but not the converter power rating. On the remaining 23 days, neither would be sufficient. On this basis Table 21.1 can be used to reach decisions about EES system size.
4. Energy storage operation
Strategies are needed to operate energy storage in a live network situation to ensure the specified control objectives are met. The required complexity can range from a predetermined schedule that remains unchanged under all network, market, and forecast conditions to a fully automated system that reads all of these and adapts under a sophisticated decision-making process. A storage system might act in isolation, in coordination with other storage systems, or in combination with other interventions, such as demand-side response, real-time thermal ratings, or onload tap changers. The fundamental energy limit of any storage media leads to the conclusion that forecasting is a key feature of any storage installation. This may be implicit in the sizing calculation that is made when specifying the system at the design stage—the operating strategy then assumes sufficient energy will be available based upon the design. Alternatively, a forecasting function may be explicitly incorporated in the operation strategy which regulates the duty to be performed by the storage system throughout its lifetime.
This section explores a number of energy storage operation strategies that work in isolation and in concert with other devices to provide a range of network control services.
4.1. Balanced and Unbalanced Power Exchange Strategies
Integrating renewable energy into LV networks brings a number of challenges to existing distribution networks, particularly steady-state voltage rise and in some cases voltage unbalance. EES systems can play an essential role in facilitating renewable energy integration by mitigating voltage rise and unbalance problems. Three-phase voltage measurements from a low-voltage network in the customer-led network revolution (CLNR) project are shown in Fig. 21.5. It can be seen that the three-phase voltages are always higher than rated voltage (1 pu, i.e., one per unit) and the maximum voltage reaches 1.075 pu, which is close to the 1.1 pu statutory voltage limit in the United Kingdom and Europe. It can also be seen that the three-phase voltages are not balanced; the maximum unbalance factor was found to be 0.45% against the limit of 1.3% in the United Kingdom.
Figure 21.5 Measured three-phase voltages from midnight to 0600.
Two EES voltage control strategies have been evaluated: balanced power exchange control strategy (BPECS) and unbalanced power exchange control strategy (UPECS). In BPECS the power flows on each of the phases of the EES are the same. In UPECS the power flow on each phase of the EES can be controlled independently. The test network for evaluating the controllers is shown in Fig. 21.6 with the domestic loads on the first branch of Feeder 1, to which a total of 15 households are connected, modeled in detail, while the loads on the remaining branches of this feeder and other feeders are modeled as aggregated load.
Figure 21.6 Low-voltage section of simulation network.
Simulation results demonstrating the operation of the EES with UPECS are shown in Fig. 21.7. The three-phase voltages at N2, which is the remote end of the feeder, are illustrated in Fig. 21.7a. The power import/export of the PV generation system and EES system are illustrated in Fig. 21.7b.
Figure 21.7 Simulation results.
It can be seen that the three-phase voltages increase after PV generation starts injecting its maximum output power into the network at time t = 5 s and the voltage of Phase C rises above the statutory voltage limit. At time t = 10 s the EES starts importing power from the network which decreases the voltage below the limit. This indicates that the EES can mitigate violation of the voltage limit effectively.
Table 21.2 shows the required EES converter power rating using each of the voltage control strategies. It can be seen that the EES converter with UPECS requires a lower three-phase power rating than that with BPECS. This result demonstrates that the ability to control power exchange independently on each of the EES phases reduces the required energy storage capacity and converter power rating. Further information on these strategies can be found in [60].
Table 21.2
Required EES Converter Power Rating
| Phase A | Phase B | Phase C | |
| EES with BPECS/kW | 11 | 11 | 11 |
| EES with UPECS/kW | 3 | 0 | 11 |
4.2. Combining Energy Storage and Demand Response
The application of demand-side response (DSR) can yield numerous network benefits, such as reduction of the generation margin and improvements to the investment and operational efficiencies of both transmission and distribution systems [58]. In [61] it was demonstrated how DSR can also be used to solve distribution network voltage problems. Trials from the CLNR project showed that DSR was not always immediately available. In one case, on receiving a call from the DNO, the customer started a diesel generator to supply power to meet their demand. The customer load was thus reduced by over 800 kW for 4 h. It is important to note that there was a delay of approximately 20 min before customer consumption was actually reduced.
Energy storage under direct control of the network operator can be called into operation very rapidly. There are examples of energy storage and DSR being used together for optimizing network capacity [62] and cost reductions [63].
Under a scenario with a high penetration of electric vehicles (EVs) and heat pumps (HPs) within a localized cluster of 230 domestic customers, the steady-state voltage limit is violated. A control scheme using EES in conjunction with DSR has been devised to solve the problem.
The demand curve with a high EV and HP scenario and the resulting voltage profile are shown in Fig. 21.8. The voltage at the remote end of the longest LV feeder drops below the statutory limit (0.94 pu in the United Kingdom) during the nighttime peak period, early morning, and afternoon peak time.
Figure 21.8 Demand and voltage profiles during 1 day.
Thick black line indicates minimum allowable voltage.
Thick black line indicates minimum allowable voltage.
To mitigate violation of steady-state voltage limits the collaborative control system will instruct the EES to operate first and export real power into the LV network to increase the voltage at the remote end. The collaborative voltage control scheme will simultaneously call for DSR. When operation of the DSR is confirmed and the steady-state voltage is within limits the collaborative voltage control system will instruct the EES to reduce real power export and thus conserve the limited storage capacity.
In Fig. 21.9 the voltage profile between 2330 and 0400 is seen to drop below the 0.94 pu limit, at approximately 0015 in the morning. The EES then injected 10 kW of real power into the grid to bring the voltage back above the limit and, at the same time, the DSR command was issued. After 20 min the consumption of the DSR customer started to reduce but did not reach a stable level until 0100. At this time the voltage of the network was close to the statutory limit, therefore the collaborative voltage control scheme decided to maintain the output of the EES to prevent a further voltage problem. However, around 45 min later, when the voltage again went below the limit, the EES started to inject more power into the network to keep the voltage above the limit.
Figure 21.9 Voltage control during early morning period.
(a) Voltage profile and (b) DSR and EES responses.
(a) Voltage profile and (b) DSR and EES responses.
It can be seen that use of the two techniques in collaboration offers benefits beyond the use of a single technique. The use of DSR on its own would result in a voltage problem sustained for approximately 20 min; this is avoided due to the fast response of the EES. The capacity of the EES required is reduced because the DSR system can remove or reduce the need for storage intervention after 20 min. Given that EES technology is currently expensive and the cost of DSR is lower than the cost of EES, this is a valuable contribution. Further information on this study can be found in [64].
4.3. Coordination of Multiple Energy Storage Units
Using more than one ESS in a coordinated control system can offer a number of potential benefits, both in the provision of the service and the operational flexibility of the storage units. To examine the relative merits of coordinating storage systems the ideas of decentralized, centralized, and coordinated control need to be examined.
In decentralized control each unit uses local measurements to control its charging/discharging function. This type of control strategy does not require a wider communication scheme and so is in some respects robust, reliable, and cost effective compared with centralized control [65]. However, due to no communication between units, support cannot be received from other units if either an extreme SOC or a power limit is reached or in the event of complete unit failure.
In centralized control the charging and discharging control actions for each unit are determined in a central controller. This approach requires online information of the network state and high computation speed [66]. A significant drawback of this control approach is cost, since it requires a fast, high-reliability communications network. In the event of communication failure each unit would not be able to respond to a voltage excursion.
In coordinated control the control strategy combines the positive features of both centralized and decentralized control [67,68]. The distinctive features of this control are robustness with respect to intermittency and latency of feedback and tolerance to connection and disconnection of network components.
A study has compared the relative strengths of these approaches to control [69]. In Fig. 21.10 battery energy storage system (BESS) decentralized controllers act on local measurements of voltage, combined with a voltage sensitivity factor, to determine active and reactive power set points. The voltage sensitivity factor (VSF) is a measure of how large an impact that variations in power will have on voltage at a given node. The centralized controller determines which BESSs should be used to solve voltage problems by considering the remaining battery cycle life, energy storage availability, and their VSF. This analysis concluded that:
• By sharing power and energy between the BESSs the scheme is able to solve real-time voltage problems that cannot be solved with independently controlled BESSs with the same power and energy capacity.
• The rated power and energy of BESS units at locations with the most severe requirements are reduced, hence the largest unit is smaller when compared with a unit in the same location with noncoordinated control.
• The even sizing of BESS units offers advantages in maintenance and economies of scale in manufacturing.
• There is greater potential for this proposed method to adapt to changes in location of extra PV generation, albeit to the limit where extra capacity would then be required. This is not the case with noncoordinated control.
• The addition of an aging model more evenly utilizes the BESSs and consequently reduces the cost of battery replacements for the storage operator, both in terms of battery replacement and maintenance requirements.
Figure 21.10 Coordinated control of multiple energy storage units.
4.4. Summary
This brief summary of storage operation studies shows how there are many considerations to be made in the provision of storage control. A single storage unit will have to provide one or more services in the areas of network control or market operations. To do so, reliance on direct or indirect forecasting of the duty to be provided is required. There can be benefits in sharing the duty required to provide a service between multiple storage units or alternative technologies such as DSR.
5. Demonstration projects
This section describes three demonstration projects on UK distribution networks. In each case, we present the ESS installed and explain what it was intended to accomplish, we present trial results and summarize the outcomes of the project.
5.1. Hemsby Energy Storage
This was the first installation of large-scale energy storage on a distribution network in Great Britain [70–72]. A picture of the storage site is shown in Fig. 21.11, and the local network, close to Great Yarmouth in the United Kingdom, is shown in Fig. 21.12. The ESS is located at a normally open point at the remote ends of Feeder 1A and Feeder 2B. A 2.25 MW windfarm is located midway along Feeder 1A. The area is not on the gas grid, so there is a significant electric heating load. Feeder 1A has peak demand (2.4 MW) occurring between 0100 and 0200, and Feeder 2B has a contrasting daytime peak (4.6 MW) when holiday parks have high occupancy.
Figure 21.12 Primary and secondary instrumentation points in demonstration network 11 kV feeder.
5.1.1. Energy Storage System
The storage system was nominally rated as a 200 kW h/200 kW network, and the storage medium selected was lithium-ion batteries. The ESS could operate in four quadrants, simultaneously exchanging real and reactive power with the network in either forward or reverse direction. The converter was rated at 850 kV A which permitted 600 kW and 600 kV Ar to be transferred simultaneously. The battery modules were able to deliver 600 kW of power, but were generally limited to 200 kW (1 C).
5.1.2. Automatic Control
To carry out a substantial series of trials, it was necessary to operate the ESS under automatic control. The control algorithm developed could read required variables, process them in a flexible manner, and issue set points for implementation on the ESS. The core control algorithm could be configured to respond to any of the values from instrumentation points on the network; this was called an “event,” with associated “location,” “threshold” (of the measurement), and “action” (power set points).
As an extension to the basic event, location, threshold, action process described earlier an algorithm was developed to track the time and magnitude of peak power flow occurring on the feeder. This function was required for this demonstration project because daily and seasonal variability in feeder power flow lead to either limited use or frequent saturation of the ESS capability.
Further to the event-driven algorithms described previously a scheduling system was developed to provide a facility to run repeated set point combinations on the ESS. This provided the capability to carry out sequences of tests on the ESS, designed to verify the operating parameters of the system.
5.1.3. Trial Results and Validation
The effect of the peak-shaving algorithm, described earlier, in conjunction with EES in reducing power flow on the 11 kV feeders is illustrated in Fig. 21.13. It can be seen that, due to the scale of the EES relative to the feeder demand, the effect of the EES system on the load profile is relatively small.
Figure 21.13 11 kV feeder power flows with and without peak shaving enabled (Aug. 7, 2013).
Operation of the EES and the impact that this operation has on estimated SOC of the battery is illustrated in Fig. 21.14.
Figure 21.14 Real power import/export and SOC of EES system (Aug. 7, 2013).
SOC negative overshoot is due to the SOC battery terminal voltage estimation technique adopted by the system [74]. When estimating SOC under dynamic conditions there is a significant voltage drop due to internal impedance, whereas when export power is reduced to zero, there is no current flow in the battery circuit and no voltage drop. Charging/discharging rates, battery age, state of health (SOH) [75], and environmental conditions (e.g., ambient temperature) have been shown to have significant impact on this internal impedance value [76]. Consequently, the instantaneous online SOC estimator is significantly lower at the end of discharging than the no-load measurement (as shown in Fig. 21.14).
Results measured from the network in Fig. 21.15 show the action of automatic voltage control. To make a comparison with the effect that would have been seen without the ESS in operation a load flow was carried out to simulate voltage at the point of common coupling (PCC), both with and without the ESS. For this early test, limited data acquisition was available on the network, so voltage at the PCC was simulated from measurements at the primary substation. This situation limits the clarity of the effect of automatic control, but the raising of low voltages and trimming of high voltages can be seen.
Figure 21.15 Reactive power exchanges with network to support voltage stability.
5.1.4. Outcomes
This project demonstrated automatic, algorithmic control of an ESS connected to the remote end of an 11 kV distribution feeder on a UK distribution network. The primary purpose of this project was to explore practical and technical features of planning, deploying and operating energy storage on an 11 kV distribution network.
5.2. Energy Storage in the Customer-Led Network Revolution
ESS were installed at multiple voltage levels as part of the CLNR project, which was the largest smart grid demonstration project ever conducted in the United Kingdom. Large-scale storage, comprising 5 MW h/2.5 MV A of lithium-ion cells, was located on the 6 kV Northern Power Grid distribution network at Rise Carr near Darlington in the north of England. The storage facility and a simplified circuit diagram of the local network are shown in Fig. 21.16 and Fig. 21.17. In addition to this a number of smaller ESS (100–200) kW h were installed on rural distribution networks. The aim of these installations was to demonstrate the benefits of individual ESS in coordinated energy storage responses, to test network interventions with multiple functions, and to interact with other network interventions such as demand response and real-time thermal ratings.
Figure 21.16 A picture of the storage at Rise Carr.
Figure 21.17 A simplified diagram of the local network.
5.2.1. Trial Results
The ESS at Rise Carr is primarily required to reduce the loading on the 11 kV/6 kV transformer. This transformer operates using a variable ampacity based on the local weather. Fig. 21.18 and Fig. 21.19 show power being supplied from the ESS to reduce demand to below the transformer thermal limit. Fig. 21.18 shows the current in the transformer with and without the action of the ESS, the transformer current, and the current supplied by the ESS. Fig. 21.19 shows the impact of delivering this service on the SOC of EES.
Figure 21.18 Real-power response for power flow management at Rise Carr.
Figure 21.19 EES state of charge during active power response.
Fig. 21.20 shows real power voltage control of a LV feeder-located energy storage unit at Mortimer Road (EES3, 50 kW/100 kW h). This is autonomous voltage control with upper and lower limits of 0.420 and 0.408 kV, respectively. Positive power implies real power export from the EES unit to the grid.
Figure 21.20 Real power control of voltage at Wooler St. Mary’s.
5.2.2. Outcomes
The CLNR EES demonstrations showed that EES can work at a variety of voltage levels and provide a mixture of network services in conjunction with other network interventions. The ESS was deployed on a variety of networks (representative of around 80% of the UK distribution network) and was of a variety of scales.
5.3. Smarter Network Storage
Installing EES for just one application is typically insufficient to make the capital investment worthwhile. The principal goal of smarter network storage (SNS) was to install large-scale EES for a variety of system benefits to maximize its value. The project installed what, at the time, was the largest battery in Europe—10 MW h/6 MW/7.5 MV A of lithium-ion storage, pictured in Fig. 21.21—at a primary substation (33/11 kV) in Bedfordshire (England). The primary purpose of the EES was to defer the need to invest in new network infrastructure by reducing peak demand on the network, shown in Fig. 21.22. To offset investment costs, revenue was gained from contracts to provide frequency response and short-term operating reserve, as well as being offered in a tolling contract to energy traders [55].
Figure 21.21 The site of the storage (left) and the battery racks (right).
Figure 21.22 Comparison between the conventional reinforcement option and the alternative using EES.
The aim of the project was to demonstrate that EES on this scale can serve as a network asset, while providing a return on investment. A forecasting and service scheduling system was developed to allow the ESS to participate in commercial services while maintaining available power and energy for the primary demand peak-shaving service.
5.3.1. Trial Results
At the time of writing, the SNS project is still ongoing, but only single-service trials have been carried out. The results of some of these trials are presented here, but ultimately the intention is to demonstrate a portfolio of services that are automatically scheduled and delivered and maximize the lifetime value of the ESS.
Fig. 21.23 shows the ESS carrying out its primary peak-shaving function. Peak demand is brought below the circuit thermal limit between 1815 and 1915, meaning that in the event of single-circuit outage the power supply could continue without any additional intervention. This is significant in that the ESS is providing a necessary reinforcement service, indicating that the industry has moved beyond demonstration projects and into ESS being used to solve real network problems.
Figure 21.23 Peak demand reduction using the ESS.
Fig. 21.24 shows the ESS providing a dynamic frequency response service. When system frequency deviates from the nominal value by more than 0.05 Hz the ESS automatically provides a power response proportionate to the deviation. This, combined with similar responses from other ancillary service providers, brings the frequency back within its limits. In the example shown the ESS is providing a bidirectional response, allowing it to act on both high- and low-frequency events.
Figure 21.24 Dynamic firm frequency response provided by the EES.
5.3.2. Outcomes
Many of the outcomes of SNS are yet to be realized because the project is, at the time of writing, still in early stage trials. However, the project is already demonstrating that it is possible to use ESS to solve real network problems, and to participate in ancillary service markets while doing so. Over the lifetime of the project, evidence will be created on the cost-effectiveness of this solution, as well as the suitability of existing markets and the regulatory framework to support this kind of approach.
6. Integrated modeling approach
A grid-scale ESS comprises three main components: the energy storage medium, a power electronic interface, and a high-level control algorithm which chooses how to operate the system based on internal and external measurements. The literature contains many examples of isolated modeling of individual energy storage mediums, power electronic interfaces, and control algorithms for energy storage. However, when assessing the performance of a complete ESS, the interaction between components gives rise to a range of phenomena that are difficult to quantify if studied in isolation. An integrated electrothermochemical modeling methodology seeks to address this problem directly by integrating reduced-order models of battery cell chemistry, power electronic circuits, and grid operation into a computationally efficient framework [77]. Physics-based battery and power converter models are implemented with sufficient detail to account for energy losses and track battery degradation through solid–electrolyte interphase (SEI) layer growth. This BESS model is coupled with a grid model. The same grid control objectives can be met through different battery operating set points, resulting in considerable differences in battery degradation and roundtrip efficiency.
The proposed benefits stemming from this approach are:
• to enable online mapping of asset characteristics and operating costs
• to assist in cost-effective sizing of grid-connected BESS
• to lead to more efficient and cost-effective operation of BESS
• to allow development and testing of novel control architectures in a realistic and detailed environment
• to facilitate incorporation of the model in real-time simulation platforms.
6.1. Methodology
Fig. 21.25 is a block diagram of the integrated model showing details of individual blocks as well as flows of power and information. Battery roundtrip energy efficiency is calculated over a cycle that begins and ends at the same SOC.
Figure 21.25 Block diagram of the proposed integrated model and details of individual blocks of the whole system.
6.1.1. Lithium-Ion Battery Model
Lithium-ion batteries can be modeled at varying degrees of complexity; 3D, 2D, 1D, and 0D. The 0D model, known as the single-particle model (SPM), has been adopted in this work. This has advantages in terms of low computational cost, but is typically only valid up to approximately 2C (the current required to empty the cell in half an hour) operation; network services are typically below this rate, so the model’s use is justified. The battery chemistry modeled is based on lithium–polymer cells with an operating voltage window of (4.2–2.7) V.
Battery voltage varies significantly with SOC, as shown in Fig. 21.26. Thus, selecting a suitable SOC range and designing the electronics around that voltage range is important when designing BESS.
Figure 21.26 Measured cell voltage as a function of SOC for a 4.8 A h Kokam lithium-ion battery.
6.1.2. Converter Model
In this study the two-stage topology presented in Fig. 21.27 is used; it allows significant variation in battery voltage, and hence more complete charge extraction, without compromising system efficiency.
Figure 21.27 Two-stage topology comprised of battery bank, DC to DC and DC to AC conversion stages.
The converter is represented by an “average mode,” based on semiconductor datasheet information, coupled to a piecewise linear approximation of a single 50 Hz fundamental period of the voltages and currents occurring in one leg of a three-phase bridge. Balanced three-phase waveforms are assumed during simulation, so power losses only need to be calculated for a pair of semiconductors. The model accounts for conduction and switching losses and losses in the filters of both the DC/DC and DC/AC converters.
6.1.3. Network Model and Control Algorithm
The section of network used to test the integrated model is based on a real LV network in the United Kingdom and is shown in Fig. 21.28. A scenario is examined with a 100 kW 50 kW h BESS connected directly to the secondary substation, as shown in Fig. 21.28, rated at 315 kV A. At present the loading limit of the transformer is not exceeded, but loading may increase such that network reinforcement is required. Under this scenario the BESS flattens the high evening load peaks and reduces transformer loading below the thermal thresholds while also providing ancillary services.
Figure 21.28 The test network used in this study, based on a real LV network in the United Kingdom.
The BESS connection to the grid uses current sources responding to active and reactive power set points. Power set points along with the voltage at the PCC are passed to the converter model, which calculates battery current and outputs this to the battery model, which calculates battery parameters.
6.1.3.1. BESS control algorithm
The control algorithm’s primary function is to follow the transformer load and respond by flattening it by discharging the battery when demand exceeds 300 kW. The algorithm issues power commands that equal the difference between the load and the 300 kW threshhold. The algorithm also accounts for SOC and current limitations; these are calculated online in the battery model and aim to maintain individual cell voltage between (2.7–4.2) V limits. The BESS will normally float at a preselected SOC; it will be allowed to charge, either prior to the evening peak or during the evening.
Fig. 21.29 displays the typical operation of the BESS, set to float at 30% SOC.
Figure 21.29 Typical operation of the BESS, showing transformer loading and BESS power.
6.2. Results and Discussions
The system was simulated, using the integrated modeling approach, for an equivalent time window of 24 h. An autumn load profile, scaled such that the evening peak exceeds the transformer loading limit, was used in this study.
Fig. 21.30 shows BESS operation for two different combinations of floating SOC as well as SOC limits:
1. Floating SOC 100%, maximum SOC 100%, minimum SOC 20%.
2. Floating SOC 30%, maximum SOC 80%, minimum SOC 0%.
Figure 21.30 BESS operation for two combinations of floating SOC and SOC limits, showing (a) power flowing through the transformer; (b) BESS power; (c) battery SOC; and (d) battery capacity loss.
Fig. 21.30 shows the power flowing through the transformer, the BESS power, SOC, battery capacity loss, and energy absorbed and released during operation of the BESS. As can be seen in Fig. 21.30d, capacity loss varies significantly depending on the floating SOC, with lower SOCs resulting in lower degradation.
The energy absorbed and released during operation of the BESS is depicted in Fig. 21.31a for combination 1. The energy dissipated in the batteries through internal irreversible losses is 33.4% of total losses. Fig. 21.31b shows the dependency of battery efficiency on the depth of discharge. Battery efficiency decreases as the depth of discharge is increased for cycles with equal SOC set points. In the case of cycles with similar depths of discharge, efficiency is higher in the cycles with the highest SOC set points.
Figure 21.31 Roundtrip energy losses and battery efficiency: (a) energy absorbed and released during operation of the BESS, and calculated total roundtrip energy losses ∆EESS and battery roundtrip losses ∆Ebat, respectively, over time for a 24 h cycle; and (b) SOC over time for a 24 h cycle marking battery roundtrip efficiency calculated for different depths of discharge and SOC values.
Simulations were then undertaken to study the effects of floating SOC and SOC limits on degradation and efficiency. The scenarios considered are shown in Table 21.3.
Table 21.3
Combinations of Different Floating SOC and SOC Limits
| Max.–Min. SOC/% | ||||
| Float SOC/% | Max. 100–Min. 20 | Max. 90–Min. 10 | Max. 80–Min. 0 | |
| 100 | V | X | X | |
| 90 | V | V | X | |
| 60 | V | V | V | |
| 30 | V | V | V | |
V, valid combination; X, nonvalid combination.
Results from these simulations are shown in Fig. 21.32. Floating at a higher SOC results in a higher rate of capacity loss. A 100% increase in capacity loss can be noted for the two extreme cases, that is, floating at 100% and 30% SOC.
Figure 21.32 Battery degradation for 24 h of operation for different combinations of floating SOC and maximum SOC limits: (a) capacity loss over time and (b) total capacity loss at the end of the 24 h operation for a 4.8 A h cell.
Fig. 21.32 shows energy losses for the same combinations. Converter losses are not greatly affected by the floating SOC, depending instead on system utilization. Fig. 21.33 reaffirms that the main contributor to losses in the BESS is the battery interface and that higher depths of discharge will result in reduced battery efficiency.
Figure 21.33 Energy losses of the ESS as the sum of battery losses and converter losses.
The efficiency of the converter and the battery is calculated based on total energy processed during the day. Energy losses are presented for different combinations of floating SOC and maximum SOC limits.
The efficiency of the converter and the battery is calculated based on total energy processed during the day. Energy losses are presented for different combinations of floating SOC and maximum SOC limits.
These results imply that the choice of both the floating SOC and the SOC limits are critical in terms of efficiency and degradation. Even though the same energy was delivered to the grid under the same power by several combinations, capacity loss can be significantly different depending on the load profile and SOC restrictions.
The previous conclusions suggest that accurate load prediction could lead to optimized BESS operation, where load demand can be met and capacity loss minimized. The results also suggest that the targets of increasing roundtrip efficiency and minimizing battery degradation are contradictory in terms of establishing SOC set points. However, these vary nonlinearly and thus require models of sufficient fidelity to capture this. It would not have been possible to establish these conclusions without adopting such an integrated modeling approach.
The model incorporates accurate calculations of the key battery and power converter properties that impose important constraints on the services provided to the grid. The latter can be taken into consideration in control algorithms targeting objectives that can range from voltage control to arbitrage and renewable energy time shift. This methodology could be applied to any other energy storage technology and thus acts as a platform for future work in this area.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.