Distribution Feeder Upgrade Deferral Through use of Energy Storage Systems

Transcription

1 1 Distribution Feeder Upgrade Deferral Through use of Energy Storage Systems Tan Zhang, Student Member, IEEE, Alexander E. Emanuel, Life Fellow, IEEE and John. A. Orr, Life Fellow, IEEE Abstract A method for determination of the optimal deployment of battery energy storage systems for the purpose of deferral of feeder capacity upgrades is developed, together with determination of the economic consequences of such deferral. This work resulted in a practical software planning platform for utility engineers. Through this work it was learned that there may be an optimal deferral period for a given situation which yields the maximum benefit. This study also reveals that the annual rate of load increase and the cost per unit length for the feeder upgrade have critical impacts on the benefits. Under a low rate of load increase with a high cost of feeder upgrade, the deferral economic benefits may be substantial. Index Terms Battery Energy Storage System (BESS), feeder upgrade deferral, peak shaving, distribution system planning. I. INTRODUCTION HE capacity of transmission and distribution systems must Texpand at a rate dictated by demographic and economic factors [1]. Utility distribution planning engineers must study the load growth forecast and plan for feeder upgrades or new installations in order to avoid exceeding the original thermal ampacity limitations of the feeder line conductor or associated transformers. Studies such as in [2] have shown that distributed generation may enable deferral of this substantial upgrade expense. A recent DOE and Sandia National Laboratories report [3] demonstrated that deferral can be accomplished by installation of properly located battery energy storage systems (BESSs) along the feeder to meet the additional peak demand. Other reports [4, 5] addressed the battery energy storage system sizing for feeder deferral in terms of power and energy rating along with the BESS total cost as the sum of the energy subsystem and power conversion subsystem costs. The purpose of this study is to develop a software platform for determining the optimal feeder deferral strategy using a BESS. The software implements a complete engineering economic study that provides the optimal ratings of the battery charge/discharge power along with its energy capacity and the feeder deferral duration which yields the maximum benefit. This is a practical approach to determination of the economic benefits provided by a BESS by incorporating the impact of the system parameters and variables such as the annual rate of load growth as well as the cost of the feeder upgrade. II. PREPARATION OF SYSTEM DATA The economic viability of a BESS for feeder deferral is a function of several parameters and variables. This work presents an organized method for evaluating the combined impacts of those quantities on cost savings. The electrical and economic quantities that govern energy storage and discharge operation are analyzed below. A. Feeder Apparent Power Limitation and the Load Profile. In a radial distribution system the power is delivered from a substation to the end users through dedicated feeders. Each feeder has a recommended apparent power limitation. This limit is dictated mainly by the feeder conductor size and allowable sag. Feeder upgrade is required when the demand exceeds feeder ampacity or the sagging of overhead conductors reduces the clearance below the minimum required value. A different upgrade situtation would be caused by the feeder load exceeding the transformer kva rating. A BESS could potentially permit transformer upgrade deferral. However, this paper addresses only the case of feeder deferral. Feeder upgrade planning is driven by projections of the magnitude and duration of peak loads, which typically follow daily, weekly, and seasonal pattern as illustrated by the actual data in Fig. 1. Each feeder exhibits patterns which in aggregate make up the load of a particular area as shown for the state of Missouri in Fig. 1. Financial support for this work was provided by Vionx Energy (Premium Power Corporation), National Grid, and the US department of Energy Tan Zhang, Alexander E. Emanuel and John. A. Orr are with Worcester Polytechnic Institute (WPI), Department of Electrical and Computer Engineering, Worcester, MA 01609, USA. For correspondence tzhang2@wpi.edu. Figure 1. Actual daily load profiles in the state of Missouri [6] The regional load profile of Fig. 1 is produced as a result of the individual demand curves of each feeder. Simulated data produced probabilistically for single feeder is shown in Fig. 2.

2 2 Figure 2. Simulated annual load variation, note the demand exceeds the feeder thermal limitation during the summer and winter peak demand days. Annual load profiles typically peak during the summer and winter months. The peak demand of the year is denoted as. With a given present load curve such as in Fig. 2 and a projected rate of load growth of (%/year) the utility engineer must plan for an upgrade before exceeds the feeder rating. This process is explained in more details under section III. B. Cost Function for Battery Energy Storage System. The energy storage system capital cost can be divided into two main parts. The first one is the cost of the power conditioning system and its auxiliaries denoted as the power component with unit price, inm$/mw. The other one is the energy component, representing the cost of the actual storage components with unit price, in M$/MWh [3]. There are various types of storage technologies. One of the most promising is the lithium-ion family of batteries. The price information from vendors along with smoothly fitted functions are presented in Fig. 3. As shown, the prices for both the power subsystem (M$/MW) and the energy subsystem (M$/MWh) are nonlinear and can be fitted into the following two expressions as functions of the studied storage system s power and energy ratings, respectively: is the power rating of the battery energy storage system in MW. Note that for this study is assumed to equal. is the energy rating of the battery energy storage system in MWh or MVAh. For example, with a rated power and discharge duration,. The capital cost of a battery energy storage system ( ) is: An important consideration is that in general the battery will degrade through its life span ( ), possibly necessitating one or more quick battery replacements during the period of deferral. The costs are highly dependent on the BESS technologies and are considered in [3] with results reported as a battery replacement factor ( ) ranging from 0 to 40% of the total cost of the original system. For example, if a capital cost of a specific battery energy storage system with rated power 1 MW, discharging duration 2 hours is 2M$, then a 20% battery replacement factor means the battery replacement cost will be 0.4M$. Figure 3. Lithium-ion family battery cost information: (a) power component price in (M$/MW) (b) energy component price in (M$/MWh) [3] Another essential part of the total operational cost is the stream of annual operating expenses in $/year. This is the sum of two main O&M cost, namely, fixed O&M in $/MW-yr and variable O&M in $/MWh for both charge and discharge. The calculations are trivial. For instance, the fixed O&M is assumed to be $9000/MW-yr, and variable O&M to be $5/MWh for charging/discharging. A rated battery energy storage system capacity of 1 MW with 1 hour discharging duration, 365 cycles per year and 85% round trip AC/AC efficiency will have annual fixed O&M of $9,000 and variable O&M of $3,972. Thus /year for this type of battery energy storage system. It is reasonable to assume that the BESS will achieve its maximum potential by addressing possible multiple use cases such as stability enhancement, renewable energy output smoothing and outage avoidance through its life span. III. BENEFITS EVALUATION METHODOLOGY Two possible trends of the growth in maximum apparent power demand supplied by a radial feeder are presented in Fig. 4, corresponding to two different annual rates of load increase, and in %/year. In both cases (as well as all others later), the maximum apparent power demand will reach the feeder capacity limitation (10 MVA in this study) at the year, is the maximum apparent power demand that can be withstood by the feeder. Prior to, the utility planner must determine, the project design and installation lead time and start the feeder construction at the time.

3 3 Maximum Load Demand (MVA) t Time (year) a t F Figure 4. Maximum apparent power demand and vs. time with showing the feeder thermal capacity limitation. This study considers a BESS with the rated charging/discharging power (when the power factor ) and rated discharging duration under. The BESS enables deferral of feeder construction by shifting the peak demand. This concept is presented in Fig. 5 and 6. Maximum Load Demand (MVA) S L S ' a S a S L S a t CF S max S max S ' max t t a F t ' Time (year) F Figure 5. Maximum apparent power demand vs. time with a BESS installation. The BESS is initially installed at. If it has enough capacity ( ) to level out the entire peak demand by charging energy during the period when the demand is low (Fig. 6), the new feeder does not need to be ready until. This non-wire solution will achieve years of deferral. The maximum deferral time can be determined by the time when the following inequality is no longer satisfied: This simply means that the stored energy must be sufficient to supply the total energy needs during the peak as illustrated in Fig. 6 which shows an excessive peak demand. Figure hour highest load demand curve of the year and indicating BESS charging energy ( ) and discharging energy ( ). t P S B ' The capital cost for feeder upgrade is denoted as and it is a function of the upgraded feeder length: = price of feeder upgrade ($/mi) = upgraded feeder length (mi) The study assumes that for a given a capital cost of a project, the utility will takes a loan from a bank and pays off the capital cost as a series of annual loan payments. The payments start when the project begins and stop when the loan duration ends. This paper uses the loan interest rate = 6%/year for the examples. The annual loan payments for feeder upgrade and battery energy storage system installation projects are: : loan duration for feeder installation and battery investment, respectively. The study results are based on and years. Considering the time value of money, the investments present values at the study starting time for both the feeder upgrade and the battery energy storage system installation have the following expressions: =discount rate (assumed 10%/year). feeder and battery installation duration, respectively. It is assumed and (years). annual loan payment for feeder and battery energy storage system, respectively. Similarly, as indicated in section II B, the present value for the BESS total operational cost ( ) during its service period can also be calculated: = the number of battery replacements needed during years of deferral with battery lifetime, determined by and. = annual operating expense, in $/year. The first term in (10) is associated with the total cost of N sequential battery replacements and the second term is associated with the stream of annual operational expenses. The method to determine the economic benefit of feeder upgrade deferral is to compare the net present values of the following two possible investment approaches at the year :

4 4 1. The construction of an additional feeder at the future time when the load grows beyond the original feeder ampacity limitation, denoted as. 2. The installation and total operational cost of a BESS at the time plus the deferred time years of new feeder construction at the time, denoted as:. In Fig. 7 is provided a flowchart that shows the procedure that leads to determination of the maximum benefit and optimal deferral duration. %/year, there will be a maximum benefit of $0.72M with 4 years of deferral. If increases to 1.0 %/year, the maximum benefit drops to $0.08M and this occurs at a much shorter deferral duration ( year). Fig. 9 also illustrates that as the annual rate of load increase rises the benefit will vanish. Under the specific assumptions in Table I, there will be no benefit when %/year. System Inputs Determine the maximum deferral duration: Initially set BESS optimal sizing by the given Benefit Calculation: Figure 8. Deferral benefits (Million $) as a fuction of the deferral duration with projected rate of load growth as a parameter.? No END Yes Figure 7. Flow chart for the engineering economics calculation. IV. RESULTS An example of the implementation of this algorithm makes use of the cost information in Table I: Table I: Cost information for feeder and BESS installation Feeder Length (:mi) 6 Price ( :$/mi) BESS Figure 9. Maximum deferral benefits (Million $) as a fuction of the annual rate of load increase ( ). Fig. 10 illustrates the impacts of battery degradation on the benefit, showing different levels of battery replacement factor. It is learned that the benefit will decrease as the cost of battery replacing is increasing. For instance, an ideal battery without degradation will result a maximum deferral benefit of 0.76 M$ with 6 years of deferral, it drops to 0.66 M$ with a shorter deferral duration ( years) as the increases to 40%. Round Trip Efficiency () 85 Calendar Life ( :years) 5 Battery Replacement Factor (%) 20 Fixed O&M ($/MW-yr) Variable O&M ($/MWh) 1.5 Fig. 8 shows the quantified benefits as a function of the deferral duration with different annual rates of load increase as a parameter. It is learned that for smaller than a certain critical value, there is an optimal deferral duration that yields a maximum benefit. For example, with Figure 10. Deferral benefits (Million $) as a fuction of the deferral duration with battery replacement factor as a parameter,, 5 years.

5 5 Four curves are plotted for different feeder upgrade prices, ranging from 0.5 to 2.0 M$/mi in Fig. 11 and Fig. 12 shows the maximum benefits for the same assumptions. These benefits range from 0.1 M$ to 2.8 M$. One learns that the cost of the feeder upgrade has a significent impact on the benefits and it is economically more attractive to defer the upgrade of an expensive feeder. for the BESS energy subsystem has a low unit price regardless of its size (0.25 M$/MVAh). The battery also has a longer calender life (10 years). For the worst case analysis, the battery replacement factor is set to be 40%. Comparing to the similar set of curves in Fig. 8, one can see the maximum benefits for the case of this specific battery are significiant with a value of $1.9M with an. Because of the cheap energy subsystem, the benefit can still be positive even under a high annual rate of load increase (i.e., 0.17 M$ with ). Figure 11. Deferral benefits (Million $) as a fuction of the deferral duration with feeder price as a parameter,,. Figure 13. Deferral benefits (Million $) for a specific low cost battery in a fuction of the deferral duration with as a parameter. Figure 12. Maximum benefits (Million $) as a fuction of the feeder upgrade price. Finally, the another simulation is performed using data for a recently-announced commercial battery storage system, Table II [7]. Table II: Cost information for a real battery energy storage system case study Feeder Length (:mi) 6 Price ( :$/mi) BESS Round Trip Efficiency () 85 Calendar Life ( :years) 10 Battery Replacement Factor (%) 40 Energy subsystem Price (M$/MVAh) 0.25 Fixed O&M ($/MW-yr) Variable O&M ($/MWh) 1.5 The set of results for the deferral benefits in this case are shown in Fig. 13. It is noted that instead of using (3), the price V. CONCLUSION This study addresses the feasibility of deferral of feeder upgrades. Not surprisingly, the strongest impacts on deferral benefits are the costs of the storage battery and the feeder upgrade. Other factors that affect the deferral benefit are the rate of load increase,, and the additional operational costs associated with the battery. For example, if the upgrading cost exceeds 1.5 M$/mi, and the annual rate of load increase is less than 0.6%/year, the benefit may exceed 1.5 M$ for seven year deferral. Moreover, availability of a less expensive BESS, with a longer useful life span will translate to a higher magnitude of benefits from energy storage system. For instance, if with 1M$/mi upgrading cost for a 6 mi feeder, a lithium-ion battery with a cost of 0.25 M$/MVAh versus 1.10 M$/MVAh, results in 1.90 M$ benefit versus 0.72 M$ benefit. ACKNOWLEDGMENT The authors gratefully acknowledge the contributions of Clayton Burns and Justin Woodard of National Grid for their assistance in accessing feeder data. We also thank Mr. Stephen Rourke and ISO New England staff for their insight into the various aspects of the cost of power. REFERENCES [1] Power Distribution Planning Reference Book, Second Edition by H. Lee Willis, ISBN-13: ISBN-10: , [2] Gil, H.A.; Joos, G., "On the Quantification of the Network Capacity Deferral Value of Distributed Generation," Power Systems, IEEE Transactions on, vol.21, no.4, pp.1592,1599, Nov [3] Abbas A. Akhil etc. DOE/EPRI Electricity Storage Handbook in Collaboration with NRECA, SAND , February Available: [4] Jim Eyer. Electric Utility Transmission and Distribution Upgrade Deferral Benefits from Modular Electricity Storage, SAND Available: [5] Abbas Akhil, Shiva Swaminathan, Rajat K. Sen. Cost Analysis of Energy Storage Systems for Electric Utility Applications, SAND Available: [6] Zaborsky and J. W. Rittenhouse, Electric Power Transmission, Rensselaer Bookstore, 1969, p [7]