Publications

X. Sun, R. Newnham, T. Gray , D. Karner, and J. Francfort, Development and Testing of an UltraBattery-Equipped Honda Civic Hybrid, INL/EXT-12-28003, August, 2012.

Abstract: In September, 2009 Electric Transportation Engineering Corporation (ETEC) was awarded Contract 82628 for the UltraBattery Development and Testing project[1] by the Idaho National Laboratory (INL), operated by Battelle Energy Alliance, LLC (BEA). The primary objective of the project authorized under this contract was to evaluate the performance of the UltraBattery® – an advanced lead-carbon technology manufactured in the United States by East Penn Manufacturing and operating in mild hybrid electric vehicle (HEV) duty. Contract 82628 was amended from time to time and ultimately encompasses ten work tasks. Task 1 through Task 3 consisted of extensive laboratory testing to characterize/benchmark and evaluate UltraBattery modules manufactured by Furakawa Battery Co. Ltd., Japan (FUB), UltraBattery modules manufactured by EastPenn Manufacturing (EPUB) and the nickel-metal hydride (NiMH) modules installed by the OEM in a 2010 Honda Civic hybrid. Task 4 through Task 8 consisted of replacing the original NiMH battery in a new 2010 Honda Civic Hybrid with a battery pack of 14 EPUB modules. After completing the initial conversion, ETEC tested the UltraBattery equipped Honda Civic in and around Phoenix, AZ under the Advanced Vehicle Testing Activity (AVTA) of the U.S. Department of Energy (DOE) FreedomCAR and Vehicle Technologies Program. Task 9 was initiated to test battery modules produced by Exide and was unrelated to the UltraBattery work conducted under Tasks 1 – 8. Finally, Task 10 was initiated to conduct additional laboratory testing of the UltraBattery as described in this report.Work was completed by ETEC on Tasks 1-9 and that portion of Contract 82628 closed out. Work on Task 10 continued through January, 2013 when the Principal Investigators for this work moved from ETEC to Electric Applications Incorporated (EAI), where work continued on Task 10.

PHEV Grid Interaction Evaluation

February 2012

The PHEV Grid Interaction Evaluation project began with conducting a demand and energy study with a converted plug-in hybrid electric vehicle (PHEV) under bidirectional charging condition at ECOtality North America (ECOtality) in Phoenix, AZ. The bidirectional charging condition had the capability of direct current (DC) fast charging and discharging. The impact of this DC fast charging and discharging on the charging building energy demand profiles as well as on its vehicle energy storage system (ESS)’s life were evaluated. The analysis of the potential advantage of utilizing the vehicle ESS to facilitate the economics of pricing differences between on- and off-peak hours and further as energy storage linked to a smart grid was included as well.To further demonstrate the PHEV and electric vehicle (EV) bidirectional electricity exchange potential for reducing the electrical power demand of the charging facility, ECOtality expanded a previous Idaho National Laboratory (INL) analysis[2] on PHEV demand and energy cost study by introducing a bidirectional DC fast charging and discharging profile to a converted PHEV (Appendix A). The use of delayed charging as well as vehicle-to-grid (V2G) and vehicle-to-building (V2B) operations were also evaluated from the aspects of three independent identities: utility, business building and vehicle owner. The study indicated a V2G scheme could be beneficial to the utility, especially if it can redistribute and re-sell the electricity saved by the business site to other end users. Other potential benefits come from the value in reduction of peak energy demand, potentially foregoing the need to turn on a peaking plant or even build one at all, and the ability to store intermittent renewable energy. It is of only marginal value to the business owner in the current cost scheme. Conversely, V2B is very favorable to the business owner because the direct electrical costs are lowered in a controllable manner. However, it would only be especially attractive to a utility if the difference between on- and off-peak rates were very large, but would be useful in allowing peaking plant electricity to be used elsewhere. For both scenarios, the vehicle owner would require compensation ($0.59 per kilowatt hour (kWh)) in order not to lose money on the proposition (because of ESS replacement). The study recommended the vehicle owner must have ultimate control over the usage of the vehicle battery pack.An A123/Hymotion ESS was used in the conversion of a MY2004 Toyota Prius, after the vehicle completed 160,000 miles fleet testing in the US Department of Energy (DOE) Advanced Vehicle Testing Activity (AVTA) program, from a traditional hybrid electric vehicle (HEV) to a PHEV. The conversion included configuring the converted vehicle to be fast charge-capable using a bidirectional fast charger and an automatic charging control system. The bidirectional vehicle fast charging events and battery performance were monitored and recorded through a GridPoint data logging system. Appendix B details the hardware and software infrastructure needed. The on road test of the converted Hymotion Prius PHEV was conducted following the DOE’s AVTA PHEV Accelerated Test Procedure[3] with DC fast charging and discharging conditions for a total of 1,000 miles. The test produced data on the effectiveness of bidirectional charging and DC fast charging in reducing electrical power demand and energy costs at the charging building. To the extent practical, interface with “smart metering” was demonstrated. Proper operation of data logger and controls was also validated.Two laboratory-based test cycles were developed to evaluate the impact of DC fast charging and discharging operation on the performance of PHEV ESSs. The test cycles were designed to simulate the demands on the conversion ESSs solely. The first “Hymotion” test cycle simulated bidirectional fast chargeable PHEV operation of a Hymotion PHEV Prius that was converted using a 4.7 kWh A123/Hymotion lithium-ion (Li-ion) ESS. The A123/Hymotion ESS (supplement pack) was operated in addition to the original vehicle Nickel-Metal Hydride (NiMH) ESS. It provided propulsion power that allowed the vehicle to travel in all-electric mode (at moderate speeds and accelerations) up to 15 miles in a charge-depleting (CD) mode but was not recharged at all during vehicle operation. The second “Energy CS” test cycle simulated the operation of an Energy CS PHEV Prius vehicle in which the standard NiMH ESS was replaced with a 10 kWh, Li-ion ESS (replacement pack) from Energy CS. It provided power for acceleration and accepted power from regeneration in a similar manner to the original NiMH unit, and can provide the additional power required to achieve a CD range of 30 miles (again, at moderate speeds and acceleration).

Two Version 2 (V2) 4.7 kWh, A123/Hymotion ESSs were purchased and cycled in the laboratory for 6 months under each test cycle. Note: as Energy CS was not able to provide a 10 kWh ESS, a 4.7 kWh Hymotion ESS was used for the Energy CS simulation, but the Energy CS test cycle was re-scaled accordingly.

An A123/Hymotion ESS pack was cycled under the Hymotion test cycle for six months at an ambient temperature of 24°C (75°F). During initial operation it was found that the thermal management system with the ESS was unable to maintain a suitable operating temperature. As a result, the system was provided with the Hymotion pack needs to be improved to avoid high ESS operating temperatures. With an average ESS temperature of 40°C (104°F), the results suggest that this ESS could provide a daily demand reduction service for 3.6 years, as well as 40,389 miles of simulated driving before the battery capacity drops below 80% of nominal. This equates to the delivery of 2,400, 100% DOD (Depth of Discharge) capacity turnovers.

A second A123 pack (10 kWh, equivalent) was operated under the simulated Energy CS PHEV/demand reduction duty at an ambient temperature of 24°C (75°F) for 6 months. With an average ESS temperature of 33°C (91°F), the results suggest that this ESS could provide a daily demand reduction service for 5 years, as well as 67,902 miles of simulated driving before the battery capacity drops below 80% of nominal. This equates to the delivery of 2,800, 100% DOD capacity turnovers.

In addition to the full-size ESS testing, smaller strings of the cells that make up the A123/Hymotion ESS were assembled and cycled at ambient 24°C (75°F) and 50°C (122°F) under the Hymotion profile. Cycling a three-cell string at 24°C (75°F) under the simulated Hymotion/demand reduction profile provided the equivalent of 9.4 years of demand reduction service, as well as 102,826 miles of simulated driving (6,110, 100% DOD capacity turnovers). The three-cell string, cycled at 50°C (122°F), provided only 27% of the lifetime energy delivered by the 24°C (75°F), three-cell string. A three-cell string operated under the simulated Energy CS/demand reduction profile at an ambient temperature of 50°C (122°F)obtained a similar result (1,100, 100% DOD capacity turnovers) to that recorded for the cells cycled at 50°C (122°F)under the Hymotion conditions (1,650, 100% capacity turnovers). The extent of this reduction is of significant concern and highlights the importance of having an effective thermal management system for Li-ion powered HEVs, EVs, and PHEVs.

Storing the evaluated Li-ion cells at elevated temperatures raised the same concern as for battery performance under elevated temperatures. Storage at 52°C (126°F) for four months at 50 and 100% SOC resulted in a capacity loss of 10% (2.5% per month). Under Hymotion test cycle conditions at 24°C (75°F) for four months, the capacity loss was only 0.72% (0.18% per month). Obviously, high-temperature storage can be very detrimental to the performance of Li-ion batteries. This result again demonstrates the importance of adequate thermal management with Li-ion battery systems in PHEVs, HEVs and EVs.

The tear-down analysis at the Commonwealth Scientific and Industrial Research Organization (CSIRO), in Australia, of failed Li-ion cells operated under simulated Hymotion duty at 50°C (122°F), suggested that the carbon anode was responsible for the cell failure. Thecathodes from the failed cells were in an excellent and equal condition to that of the new cells. If this behavior is ‘across the board’ for Li-ion technologies, it is obvious that improved anode materials are required to achieve acceptable high-temperature performance.

Finally, the dynamometer fuel economy and emissions testing was conducted on the converted Hymotion Prius at the Advanced Powertrain Test Facility (APTF), of the Argonne National Laboratory (ANL), using the AVTA PHEV test procedures[4]. The converted vehicle gave an average of 155.2 miles per gallon (mpg) fuel economy with air conditioning (A/C) off and 112.9 mpg fuel economy with A/C on when driven in CD mode with a cold start. With a hot start, in CD mode, the vehicle delivered 170.3 mpg fuel economy with A/C off and 151.7 mpg with A/C on. Under charge sustaining (CS) mode, the vehicle average fuel economy reached 53.6 mpg with A/C off and 43.4 mpg with A/C on.

[1] Refer to: X.Sun, et al. Development and Testing of an UltraBattery-Equipped Honda Civic Hybrid http://avt.inl.gov/pdf/hev/UltraBatteryReport.pdf

[2] A. Masters, J. Wishart and J. Francfort, Tacoma Power/AVTA PHEV Demand and Energy Cost Demonstration Analysis Report, INL/EXT-10-18207, May, 2010.

[3] D. Karner, R. Brayer, D. Peterson, M. Kirkpatrick and J. Francfort, AVTA-PHATP01 – Plug-in Hybrid Electric Vehicle Accelerated Test Procedure, Plug-in Hybrid Electric Vehicle Integrated Test Plan and Evaluation Program, p360, INL/EXT-01-12335, March, 2007.

[4] D. Karner, R. Brayer, D. Peterson, M. Kirkpatrick and J. Francfort, AVTA-PHTP07 – Implementation of SAE J1634 May1993 – “Electric Vehicle Energy Consumption and Range Test Procedure”, Plug-in Hybrid Electric Vehicle Integrated Test Plan and Evaluation Program, p275, INL/EXT-01-12335, March, 2007

Development and testing of an Ultrabattery equipped Honda Civic

January 2012

The UltraBattery retrofit project DP1.8 and Carbon Enriched project C3, performed by ECOtality North America (ECOtality) and funded by the U.S. Department of Energy (DOE) and the Advanced Lead Acid Battery Consortium (ALABC), are to demonstrate the suitability of advanced lead battery technology in Hybrid Electrical Vehicles (HEVs).A profile, termed the ‘Simulated Honda Civic HEV Profile’ (SHCHEVP) has been developed in project DP1.8 in order to provide reproducible laboratory evaluations of different battery types under ‘real-time’ HEV conditions. The cycle is based on the Urban Dynamometer Driving Schedule (UDDS) and Highway Fuel Economy Test (HWFET) cycles and simulates operation of a battery pack in a Honda Civic HEV. One pass through the SHCHEVP takes 2,140 seconds and simulates 17.7 miles of driving. A complete NiMH battery pack was removed from a Honda Civic HEV and operated under the SHCHEVP to validate the profile. The voltage behavior and energy balance of the battery during this operation was virtually the same as that displayed by the battery when in the Honda Civic operating on the dynamometer under the UDDS and HWFET cycles, thus confirming the efficacy of the simulated profile.An important objective of the project has been to benchmark the performance of the UltraBatteries from both Furukawa Battery Co., Ltd., Japan (Furakawa) and East Penn Manufacturing Co., Inc. (East Penn). Accordingly, UltraBattery packs from both Furakawa and East Penn have been characterized under a range of conditions. Resistance measurements and capacity tests at various rates show that both battery types are very similar in performance. Both technologies, as well as a standard lead-acid module (included for baseline data), were evaluated under a simple HEV screening test. Both Furakawa and East Penn UltraBattery packs operated for over 32,000 HEV cycles with minimal loss in performance, whereas the standard lead-acid unit experienced significant degradation after only 6,273 cycles. The high-carbon, ALABC battery manufactured in project C3, was also tested under the advanced HEV schedule. Its performance was significantly better than the standard lead-acid unit, but was still inferior compared with the UltraBattery. The batteries supplied by Exide as part of the C3 project performed well under the HEV screening test, especially at high temperatures. The results suggest that higher operating temperatures may improve the performance of lead-acid based technologies operated under HEV conditions – it is recommended that life studies be conducted on these technologies under such conditions.Individual Furakawa UltraBatteries have been operated according to the SHCHEVP under a range of state of charge (SOC) windows and temperatures. Battery cycling was conducted using three different SOC windows (43-53%, 53-63%, and 63-73%) and three different battery temperatures (10oC (50oF), 30oC (86oF), and 58oC (136oF)). The results suggest that an adequate compromise between vehicle acceleration and charging efficiency during regenerative braking is provided with a SOC window of 53-63% SOC. Also, low operating temperatures severely decrease the energy returned by simulated regenerative braking. At 30oC (86oF), the number of simulated vehicle miles covered before a simulated engine recharge is required is 142 miles; the number of miles drops to less than 18 miles at 10oC (50oF). As a result, operation in cooler climates where trip distances are short (i.e., where there is insufficient time for batteries to heat up) will result in increased fuel usage. The lower temperatures also decrease the available discharge power, although this change is small relative to the effect on the charging efficiency.In another test, an individual 12 volt (V) East Penn UltraBattery was cycled for 167,700 simulated vehicle miles under the SHCHEVP (at 30oC (86oF)). While the discharge capacity decreased from 7.6 to 4.5 Ah, the battery was still capable of providing the power required for acceleration. Also, the battery’s ability to accept energy from regenerative braking decreased significantly during the operating period. The effect of this behavior on fuel economy, however, is not known. This aside, the result is considered very promising, since the SHCHEVP used to cycle the battery has the same discharge/charge intensity and frequency that is used for the NiMH battery currently in the Honda Civic Hybrid (i.e., the power levels were not decreased for the UltraBattery). This result demonstrates that the UltraBattery packs can last the design life of modern HEVs.

A 12V, NiMH module (from the Honda Civic vehicle) was tested for almost 80,000 simulated vehicle miles under the SHCHEVP (at 30oC (86oF)), and its capacity and performance remained unchanged during the test period. It consistently delivered 159 simulated vehicle miles between simulated engine recharges. The performance of the NiMH module also decreases when the temperature is lowered, although this drop is not as severe as for the UltraBattery. For example, at 10oC (50oF), the NiMH battery is still capable of operating for 71 simulated vehicle miles between simulated engine recharges, compared to only 18 miles for the UltraBattery. Therefore, the fuel usage at low temperatures of a NiMH-based HEV is expected to be lower than that of an equivalent UltraBattery-powered HEV. The extent of such a difference, however, is not known. An individual 12V, high-carbon ALABC module was also operated under the SHCHEVP, but failed after providing 40,391 miles of simulated service.

A Furakawa UltraBattery pack operated trouble-free for 60,000 simulated miles under the SHCHEVP (at 30oC (86oF)) with a minimal drop in performance. A vehicle-sized pack of East Penn UltraBattery packs also delivered 60,000 miles under the SHCHEVP (at 30oC (86oF)). While there was an initial battery failure in this pack (at 10,000 miles), logging of individual 12V modules has shown that all units were still very close in performance at the end of the cycling period. These results are very promising, and combined with the results for the individual module cycling, suggest that UltraBattery packs may be capable of lasting the design life of a modern HEV (e.g., 160,000 miles). In the C3 project, a vehicle-sized pack of the high-carbon ALABC modules was operated under the SHCHEVP, although it failed after just 27,000 simulated miles. A vehicle-sized, high-carbon, lead-acid battery from Exide was also cycled under the SHCHEVP, but it failed after just 12,500 simulated miles.

The project DP1.8 also consists of a retrofit of the original NiMH battery with a pack of 14 Ultra-Battery modules, manufactured by East Penn, in a new 2010 Honda Civic HEV. After completing the initial conversion, ECOtality tested the HEV in accordance with, and in cooperation with, Advanced Vehicle Testing Activity (AVTA) of DOEs FreedomCAR and Vehicle Technologies Program.

ECOtality conducted a full vehicle baseline characterization on the converted HEV. A full dynamometer evaluation (e.g., measurement of fuel economy under standard driving schedules on the dynamometer) was completed by Argonne National Laboratory (ANL). This approach allowed direct performance comparisons with the UltraBattery against the technologies used in the unaltered HEVs.

In October, 2011, the converted HEV was put into ECOtality’s fleet of test vehicles in Phoenix, Arizona, and it currently still being tested. The converted HEV accumulates approximately 5,000 miles on a monthly basis and is experiencing a wide range of driving conditions. The monthly data being collected from the vehicle is an array of battery parameters, such as:

  • most restrictive temperature
  • pack voltage
  • power
  • vehicle parameters, such as speed

The individual module voltages and cell/module voltage deviation are being measured separately on a monthly basis, as well as monitoring the health of individual battery modules. The mileage driven and gallons of gasoline used monthly are being recorded to monitor the vehicle average fuel economy for the month.

Simulated performance of Li-ion batteries under electric scooter duty using resistance-free charging techniques.

May, 2009

A battery pack comprising seven, 3.5 V, 31 Ah Li-ion modules was obtained from Amita Technologies Inc. It is a typical low-cost, lithium ion-battery pack supplied for powering electric scooters. A single module from a seven module pack has been operated in the laboratory under a simulated, simple urban, electric scooter schedule. The battery delivered 53 Wh during each of the completed 2300 cycles. Operating temperature was in the range between 35 to 45 C, and the time required for charging was 13.5 min (40 to 90% SOC), which increased to 16 min towards the end of life. Based on the energy requirements of the ECE 47 electric scooter urban drive cycle, it is considered conservative to predict that a battery pack comprising seven of these modules should provide between 20 000 and 25 000 km before replacement is required. The excellent cycle life recorded is attributed in part to the resistance-free charging algorithm used in these experiments. This procedure maintains very high charge rates (thereby reducing charge times), whilst avoiding undue decomposition of the solvent and electrode materials.

Production and test of VRLA batteries designed specifically for high-rate, partial-state-of-charge operation: Simulated Silverado and electric bus duty

December 2007

Electric Applications staff have been contracted to oversee the manufacture and perform simulated testing of an advanced valve-regulated lead-acid (VRLA) battery under simulated hybrid-electric vehicle (HEV) duty and in a vehicle in the field. One hundred batteries (nominal 27 Ah, 1-h rate) were constructed in conjunction with a local battery manufacturer, and included an optimum negative paste formulation and grid design. Modified batteries were evaluated under a simple and advanced HEV schedule in the laboratory. Their performance was excellent compared to an unmodified baseline unit – modified batteries delivered almost 15,000 cycles with minimal degradation in performance, whereas the unmodified unit experienced considerable polarization during charging and reached the cut-off voltage limit after just 3,000 cycles.A modified battery was evaluated under simulated HEV bus duty – the battery operated trouble-free for 515 250 cycles.   In terms of lifetime energy, this cycle life is consistent with the completion of 3300 cycles between 0 and 100% SOC. As the voltage at the end of the major discharge (200 A) only decreased slightly from 11.64 to 11.30 V during testing, it seems likely that over 4000 cycles are likely – note that such a lifetime energy delivery equates to ~70% of that provided by a Prius nickel metal hydride battery over its 160 000 mile design life. This is an excellent performance, given that the profile was based on a 50 Ah module (i.e., the profile was not rescaled to allow for the smaller 30 Ah battery used). Hence, hybrid bus manufacturers can now significantly reduce the size, weight and cost of their battery systems by replacing 50 Ah modules with the same number of 30 Ah units.A drive cycle (termed the ‘Simulated Silverado HEV Schedule) that simulates the operation of a battery pack in a Silverado HEV has been developed. It is based on data obtained from a battery pack in a Silverado during both operation under standard drive cycles on a dynamometer, and actual operation on the road. The profile is designed to apply the same energy and power requirements to the battery as that experienced in the field. A 36-V battery pack comprising modified modules has been evaluated under this schedule for a total of 55 000 simulated miles. During this time, the change in behavior of the battery voltage during charge and discharge was minimal – analysis confirmed that the batteries were in a good condition at the completion of the test, and HPPC testing showed that the ‘all important’ available discharge power had not changed as a result of the cycling. This is considered an excellent result given the abusive nature of the simulated schedule.A 36-V battery pack of modified units has been retro-fitted to a Silverado HEV and then driven around Phoenix for a total of 30 000 miles. The battery performance was trouble-free during this time, and HPPS testing showed that available discharge power had been maintained throughout the test period. In summary, the performance of the modified batteries under simulated HEV bus duty was excellent. They also performed very well under both simulated and field Silverado HEV duty – the results suggest that the ALABC SuperC technology is now ready for.

Simulated testing of gelled-electrolyte batteries under forklift/fast charge and frequency regulation conditions.

November 2006

Recently, there has been a resurgence of interest in the performance and application of lead-acid batteries. This has been driven mainly by benefits in performance obtained by increasing the amount of carbon in the negative plates to concentrations more than 20 times higher than traditional levels. Typically the carbons used have been carbon blacks and carbon fibers, or mixtures there of. EAI staff have been contracted to evaluate the performance of a range of gelled-electrolyte batteries with activated carbon added to the negative paste.Previous work in this area has focused on battery performance under high-rate, partial-state-of-charge (PSOC) conditions, similar to that experienced by hybrid-electric vehicles (HEVs), i.e., small PSOC state-of-charge windows (~ 5%). The addition of activated carbon to both absorptive glass-mat (AGM) and gelled-electrolyte, valve-regulated lead-acid battery (VRLA) types was shown to provide an enormous improvement in performance under these conditions.It is also of great interest to establish if the addition of carbons to VRLA batteries can provide benefits in the more traditional applications, where the batteries experience heavy duty cycling (i.e., the PSOC operating window is much larger). This report describes the performance of purpose-built, gelled-electrolyte and AGM batteries with elevated carbon levels, operated under both fast-charge PSOC motive-power duty in forklifts, and in utility frequency stabilization operation. The PSOC window is much larger than that experienced in HEVs (~50% rather than ~5% state-of-charge window), however, the charge and discharge currents experienced by the batteries are lower.Six types of gelled-electrolyte batteries were purpose-built using different concentrations of carbon. Individual modules and also one full size pack comprising 12, 800 Ah modules were then cycled to failure under specially developed fork lift truck and frequency regulation schedules. The gelled-electrolyte batteries provided very good cycle lives when operated under controlled fast-charge/PSOC conditions (e.g., opportunity fast-charge, utilities and motive power operation) where PSOC windows are larger (30 to 60% of capacity) and charge and discharge rates are moderate (~ 1C or less), provided that there is adequate thermal management. The addition of extra carbon to the negative plates of gelled-electrolyte batteries, however, does not seem to provide performance benefits under this type of duty. This is in contrast to that observed when small PSOC windows are employed (i.e., HEV duty), where additional carbon provided huge benefits. The major issue appears to be related to a lowering of charging efficiency associated with the presence of additional carbon that requires additional equalization charging. Finally, it appears that AGM batteries (which performed excellently with high levels of carbon and small PSOC windows, i.e., HEV duty), with or without carbon, are clearly inferior to gelled-electrolyte batteries when the PSOC window is significantly larger, i.e., 50-60% of capacity, with or without carbon.

D. Karner and J. Francfort, US Department of Energy Hybrid Electric Vehicle Battery and Fuel Economy Testing, Journal of Power Sources, 158, 1173-1177, 2006

Abstract: The advanced vehicle testing activity (AVTA), part of the US department of Energy’s FreedomCAR and Vehicle Technologies Program, has conducted testing of advanced technology vehicle since August 1995 in support of the AVTA goal to provide benchmark data for technology modelling, and research and development programs. The AVTA has testesd over 200 advanced technology vehicles including full-size electric vehicles, urban electric vehicles, neighborhood electric vehicles, and internal combustion engine vehicles powered by hydrogen. Currently, the AVTA is conducting a significant evaluation of hybrid electric vehicles (HEVs) produced by major automotive manufacturers. The results are posted on AVTA web page maintained by the Idaho National Laboratory. Through the course of this testing, the fuel economy of HEV fleets has been monitored and analyzed to determine the ‘real world’ performance of their hybrid energy systems, particularly the battery. The initial fuel economy of these vehicles has typically been less than determined by the manufacturer an also varies significantly with environmental conditions. Nevertheless, the fuel economy had, therefore, battery performance, has remained stable over the life of a given vehicle (160 000 miles).

Performance of Vision and Lichpower gelled-electrolyte batteries under simulated electric-vehicle duty.

April 2004

Electric Applications Inc. (EAI) staff have been monitoring a fleet of electric vehicles (EVs) manufactured by the Frazer Nash (FN) company for several years. As part of a continuing development process, the suitability of two new gelled-electrolyte batteries for use in these vehicles has been evaluated. The batteries, manufactured in China by the Vision and Lichpower battery companies, have been cycled under simulated EV conditions developed specifically to mimic the operating conditions of the FN EVs.Two Vision batteries delivered 40 cycles under simulated EV service before their capacities decreased to a specified cutoff level. The performance of the batteries was recovered with an equalization procedure, but after a total of 90 cycles (which included another pass through the equalization process), both units had experienced a significant, permanent capacity loss. By contrast, the Lichpower units operated for 90 cycles before requiring equalization. These batteries also experienced a permanent loss in performance during the 90 cycles, but this decrease was less than that experienced by the Vision units. The performance of both technologies under the simulated FN duty is considered unsatisfactory as it represents only 90 days of field service.A Vision and Lichpower battery were discharged to 3 V at a low current (i.e., 1 A) in order to simulate extreme deep-discharge duty. Both batteries performed very well ¾ their capacity was still acceptable after the completion of 25 cycles.The results from this program, combined with previous studies on the Sonnenschein gelled-electrolyte battery, suggest that the EV performance of this battery is superior to that of the Vision and Lichpower designs. The inexpensive Chinese batteries, however, could still be cost-effective in selected applications.

Performance of selected valve-regulated lead-acid batteries under simulated push back tractor duty.

February 2004

Electric Applications Inc. staff were contracted to develop an electric push back tractor (EPBT) that demonstrates the advantages of using a high power density, high voltage battery and the safety features of state-of-the-art electric vehicles.In the first phase of the program, a simulated operating strategy that mimics EPBT operation at airports was developed to allow the testing and evaluation of various candidate batteries for use with the EPBT. Seven valve-regulated lead-acid (VRLA) battery technologies were then selected (based on a combination of voltage, cost, capacity and claimed partial-state-of-charge performance) for evaluation under the EPBT schedule. Each battery type was then subjected to repeated discharges that simulate EPBT duty. Only two of the battery types were found to meet the energy, power and reliability requirements that have been established for successful EPBT operation, namely, a gelled-electrolyte VRLA battery from Sonnenschein (6-V; 110 Ah, 1 h rate) and an absorptive glass-mat (AGM) VRLA battery from Vision (12-V; 124 Ah, 1 h rate).In the second phase of the work program, the Sonnenschein and Vision batteries have been operated under simulated EPBT conditions to evaluate the potential cycle-life of these batteries under field conditions. The simulated conditions are based on what is called partial-state-of-charge (PSOC) duty. This mode of operation involves cycling batteries below a full state-of-charge (SoC) for extended periods, between full recharges. Such operation has been shown to extend the lifetime Ah’s available from VRLA batteries significantly, provided that the frequency and intensity of the full recharge is carefully controlled. Two Sonnenschein units have performed almost 6000 simulated towing operations (i.e., approximately two years of simulated field service), and are still at 90% of their nominal capacity. Based on this data, and the fact that the simulated conditions used in this study are more severe than those expected in the field, it is predicted that under normal operating conditions, the batteries should be capable of providing at least four years of field service. These results have greatly exceeded expectations and provide direct evidence that EPBTs can be less expensive than diesel-powered alternatives.The first two Vision batteries evaluated did not perform as well as the Sonnenschein units and failed before completing 1200 simulated towing operations. This poor performance is attributed mainly to the use of a charging strategy that was not suitably optimized for the specific battery type. Another two Vision modules operating under an improved charging regime delivered 2700 and 3400 towing operations before failure. Given that the Vision batteries are effectively half the price of the Sonnenschein units, the former are also considered to be an economic option for powering EPBTs. Finally, the performance improvement obtained for the Vision batteries by optimizing the operating strategy reinforces the importance of tailoring charging algorithms to both battery type and application.

R.H. Newham and W.G.A., Advanced Management Strategies for Remote Area Power-supply Systems, 10th Asian Battery Conference, Bangkok, Thailand, September 3-5, 2003

Abstract: An operating strategy based on partial-state-of-charge (PSoC) operation has been developed for a remote-area power-supply (RAPS) system in Peru. The facility will power an entire village and comprises a photovoltaic array, a bank of gel valve-regulated lead-acid (VRLA) batteries, a diesel generator, and a sophisticated control system. The PSoC schedule involves operation below a full state-of-charge (SoC) for 28 days, followed by an equalization charge. The schedule has been evaluated by operating a 24-V battery band under simulated RAPS conditions in the laboratory. It is found that operation between 58 to 83% SoC causes the negative-plate potentials to move to significantly more negative values during charging as the PSoC duty progresses. This behavior is undesirable, because it can lead to the activation of a preset limit and a subsequent reduction in system efficiency. Lowering the PSoC window to 47-72% SoC or 40-65% SoC during the 28-day cycle is found to stabilize the negative-plate potentials. The behavior of the negative-plates in gel batteries is very similar to that observed for absorptive glass mat (AGM) designs of VRLA batteries operated in hybrid electric vehicles.

Performance of valve-regulated lead-acid batteries under simulated electric-vehicle duty.

December 2002

Electric Applications Inc. (EAI) staff have been developing infrastructure for the operation of electric vehicles (EVs) for over 10 years, including both hardware and operating strategies. In this project they developed operating strategies for EVs supplied by the Frazer Nash Company. The vehicles were fitted with VRLA batteries and charged with low-power, on-board chargers. EAI staff were employed to determine the expected cycle life of two AGM batteries (designated AGM1 and AGM2) when operated in these EVs.EAI staff developed a simulated discharge profile based on data obtained during operation of a Frazer Nash vehicle around downtown Phoenix. Three overall operating schedules have been formulated by combining this discharge profile with either of three different charging procedures. The first overall operating schedule, termed operating-schedule 1, completes the charging procedure with a constant current. The second schedule, called operating-schedule 2, finishes charging with a dI/dt control. An overall operating strategy has also been developed specifically for the AGM2 batteries (12-V; 67 Ah).Two AGM1 batteries cycled according to operating-schedule 1 performed only 60 cycles before failure, whereas two operated under operating-schedule 2 delivered 70 and 153 cycles before failure. The batteries also experienced a very high self-discharge rate. Based on the poor cyclic performance and the unfavorable self-discharge data this technology, in its current state of development, is not considered sufficiently robust for use in the Frazer Nash EVs.Two AGM2 batteries provided 215 and 220 cycles before their capacity decreased to 50% of the nominal value. As a result of this poor performance, they are not recommended for use in the Frazer Nash vehicles.

Performance of valve-regulated, lead-acid batteries under partial-state-of-charge/fast-charge and high-rate cycling conditions.

December 2002

Electric Applications Inc. staff have been developing infrastructure for the operation of electric vehicles for over 10 years. The major research focus has been the refinement of PSOC/fast-charge strategies for lead-acid battery powered EVs. They have already demonstrated that this mode of operation can provide both acceptable battery life time and extended daily vehicle range. Most of the developmental work in this area has been performed using Genesis batteries manufactured by Hawker Energy Products Inc. In the current study, batteries from Exide (Orbital), Hawker (Genesis), Electrosource (Horizon) and Optima Batteries (Optima) have been evaluated in the laboratory under both PSOC/fast-charge and high-rate discharge duty.Two Orbital and two Genesis batteries provided 580 and 640, and 680 and 730 cycles, respectively, under PSOC/fast-charge (3.6C1) duty. The water consumption of the Orbital batteries was significantly higher than that of the Genesis technology, and it is considered that this loss have affected the cycle-life of the former. It is considered that modifications to the PSOC/fast-charge algorithm, in terms of TOCV or equalization time/current, should reduce the water consumption of the Orbital battery and, therefore, improve its cycle life.Three Horizon batteries were operated under PSOC/fast-charge duty. The first two units, which had been operated in a previous EV testing program, provided 425 and 450 cycles before failure. The third battery, which was obtained directly from the Electrosource factory, performed 485 cycles. All the modules experienced a high degree of water loss during operation, and it is considered that this factor contributed significantly to their relatively poor cycle lives. As with the Orbital batteries, it is considered that modification of the PSOC/fast-charge algorithms would improve the performance of the Horizon modules. Finally, all the Horizon modules experienced significant self-discharge before cycling commenced (i.e., at least 10-15%/month). The first Optima battery operated under PSOC/fast-charge duty provided 428 cycles. On advice from Research Chemists at Optima batteries, a second unit was subjected to an additional formation procedure prior to commencing PSOC/fast-charge duty. This second battery provided 749 cycles and recorded a water loss similar to that of the Genesis batteries.The energy density of the Orbital, Genesis, Horizon and Optima battery types was also measured at rates varying from C3 to 0.2 C1, after the completion of 60 standard cycles (100% DOD, C1 rate). The energy densities obtained for the batteries was in the following order (highest to the lowest): Horizon > Optima ³ Genesis > Orbital.The cost of the lifetime Ahs delivered under PSOC/fast-charge duty has also been calculated. This is expressed in Ahs/initial cost (US$) and is based on current prices. The Genesis, Orbital, Optima and Horizon batteries provided 222, 180, 169 and 69 Ahs/initial cost, respectively. Hence, the Genesis battery is the most cost-effective option in terms of Ahs/$, although its energy density is slightly less than that of the Horizon and Optima technologies. Electrosource predict that they can reduce the cost of the Horizon batteries by almost 50% in the short to medium term. If this is achieved, the cycle life of the battery stills requires significant improvement before it can become cost-competitive.

Optimized Control of Battery Banks in Renewable Energy Systems to Mitigate Diesel-generator Emissions in the Amazon Region of Peru

June 2002

Traditionally, Peru has relied on hydroelectric systems to supply the majority of its electricity. At present, these systems account for 82% of the energy generated by the national grid system. By contrast, isolated rural communities in Peru use diesel generators as remote-area power-supply (RAPS) systems to meet their electrification needs. Whilst such generators can provide power in areas remote from the grid, they can be expensive and can produce significant quantities of carbon dioxide and related pollutants. In order to provide a more cost-effective, lower-polluting option, a program to develop and demonstrate the viability of large-scale, solar-diesel-generator-battery RAPS systems in the Amazon Region of Peru was commenced. Electric Applications Inc. staff were employed to develop advanced operating strategies that both minimize pollution from diesel-generator operation and maximise battery life, and evaluate battery performance under this duty. The schedule was based on the partial-state-of-charge (PSoC) principle, and involves operating the batteries below a full state-of-charge (SoC) for extended periods, i.e., 28-days.Two 24-V battery banks comprising thick, flat-plate gelled-electrolyte technology, termed 24V1 and 24V2, were operated under the preliminary strategy. Each bank has completed 4 months of service, i.e., four, 28-day, ‘master cycles’. Bank 24V1 commenced cycling immediately after conditioning, whereas 24V2 was first subjected to a procedure developed specifically to age the batteries. Using a method developed for estimating SoC based on open-circuit voltage (OCV) measurements, it is found that controlling charge-discharge by ‘Ah counting’, results in a 3% and 13% decrease in the SoC at the end of the master cycle for the new (24V1) and the aged (24V2) batteries, respectively. Cycling also shows that the maximum voltage of both banks attained during the daily charge period increases significantly during each master cycle. This can give rise to charging inefficiency. Given this behaviour, and the observed limitation of the Ah counting method, it is recommended that the number of PSoC cycles per master cycle be reduced to 14.It is further recommended that the SoC, as determined by Ah counting, should be adjusted on a regular basis, using the OCV-SoC relationship. The diesel generator should commence charging if the SoC decreases below 40% (as determined by Ah counting). If this occurs within 13 PSoC cycles, the generator should cease operation when the average cell voltage in the battery string reaches 2.45 V.   If the charging is activated at the end of the 14th PSoC cycle, a battery-conditioning step should be applied.

Performance of valve-regulated lead-acid batteries under electric-vehicle ground support equipment conditions. Pack cycling of Sonnenschein Gel Batteries

February 2002

Electric Applications Inc. (EAI) staff has been developing infrastructure for the operation of electric vehicles for over 10 years. They have developed partial-state-of-charge (PSOC)/fast-charge strategies for a variety of lead-acid battery powered electric vehicles (EVs). This type of duty has been shown to decrease charge times to 10-30 min, whilst retaining if not improving battery cycle-lives. Recent developments have included vehicles and operating systems for electrically powered ground-support equipment (GSE) at airports.The gelled-electrolyte valve-regulated lead-acid (VRLA) batteries currently used for GSE vehicles are expensive and have a high internal resistance. This leads to significant heating of the batteries during charging and also extended charge times (up to 1 h). To address these issues, PNW have evaluated the performance of an alternative technology manufactured by Sonnenschein that has both a reduced cost and a lower internal resistance. The batteries, called Dryfit, are a 6-V monobloc with a 1C capacity of 110 Ah.This report describes the evaluation of two GSE vehicle battery packs that comprise 28, Dryfit modules, configured in two, 84-V strings, under simulated GSE service. Both a brand new pack, and a pack retrieved from the field have been tested. The latter had been in service at the Phoenix airport for over 18 months and had performed an estimated 350 GSE cycles. The new and old packs were subjected to 204 and 24 days of simulated GSE service, respectively. The performance of both packs during these cycling periods was excellent. Indeed, the capacity provided by the old pack matched that of the new. The excellent condition of the old pack was confirmed by a series of standard 1C discharge/charge cycles ¾ the capacity obtained during these experiments was at least equal to that of the nominal value. Given the service life already provided by the old pack (see above), it is considered conservative to predict a lifetime in the field of over three years.

Performance of valve-regulated lead-acid batteries under electric-vehicle ground support equipment conditions. Final Report: Dryfit, Dominator and Energel batteries.

January 2002

Electric Applications Inc. (EAI) staff has been developing infrastructure for the operation of electric vehicles for over 10 years. They have developed partial-state-of-charge (PSOC)/fast-charge strategies for a variety of lead-acid battery powered electric vehicles (EVs). By operating vehicles in this manner, charging times of 10-30 min can be obtained whilst retaining if not improving battery cycle-lives. More recent developments have included vehicles and operating systems for electrically powered ground-support equipment (GSE) at airports.The valve-regulated lead-acid (VRLA) batteries currently used for GSE vehicles are expensive and have a high internal resistance. This leads to significant heating of the batteries during charging and extended charge times (up to 1 h). EAI staff have developed a simulated GSE operating strategy based on actual data from the field and has used this strategy to evaluate the performance of three alternative VRLA batteries that have both a lower internal resistance and cost to that previously used. The brand names of the batteries are Dryfit, Dominator and Energel and the manufacturers are Sonnenschein, East Penn and Battery Energy Power Solutions, respectively.The Dryfit batteries completed 570 GSE cycles between ~30-80% state-of-charge before their performance decreased to an acceptable level (note, one cycle is equivalent to one day of GSE service). This deterioration was attributed to the combination of a gradual decrease in charging efficiency and a charging algorithm that does not provide overcharge in an interactive manner. The net result of this behavior is that the batteries experienced a gradual ‘walk down’ in capacity as a result of insufficient charge return. It is considered that optimization of the algorithm specifically for this technology would provide a significant increase in cycle life.

The Dominator batteries delivered only 380 cycles before failure. The decrease in performance was attributed to both insufficient charge return, as well as some irreversible degradation of the battery. It is considered that fine tuning the operating algorithm for this technology would provide a minimal improvement in cycle-life.

The Energel batteries provided 1030 simulated GSE cycles before their performance dropped to an unacceptable level. This performance is considered excellent, given that the algorithm has not been fully optimized for this battery technology.

The affordability of the three technologies in terms of lifetime Ahs delivered per US$ of capital cost has been calculated. The Energel, Dryfit and Dominator batteries delivered 510, 213, and 217 Ah per US$, respectively.

Evaluation of EAI fast-charge algorithm in electric Scooter cycling duty using a Hawker Genesis absorptive-glass-mat battery.

June, 2000

In an effort to reduce air pollution, the government of Taiwan has instituted a project to develop electric motor scooters to replace the 21 million internal combustion powered motor scooters currently in use on Taiwan. As part of this project, specifications for electric scooter batteries have been issued and performance standards set for battery life under simulated electric scooter cycling conditions. Electric Applications Inc (EAI) staff have conducted laboratory cycling of a Hawker Genesis® battery meeting the standard specification for an electric scooter battery. The battery was discharged using the simulated electric scooter discharge cycle (ECE-47) and charged using a fast charge algorithm developed by EAI staff specifically for small absorptive glass mat batteries such as those specified for electric motor scooter use.A single Hawker Genesis® 26 Ah module (21.5 Ah at C1) was cycled for 9,714 simulated electric scooter discharge cycles. The battery was charged using 400 EAI fast charge cycles with an average duration of 16.1 minutes (40 to 90% SOC). The battery delivered 5,133 Ah, the equivalent of 7,771 km of electric motor scooter use. At the end of cycling, the battery capacity was 77% of its initial nominal (Ca) capacity.

R.S. Hobbs, F.A. Fleming and D.B. Karner, EV Battery Life Extension in Field Testing, SAE Technical Paper Series, 970243, International Congress & Exposition, Detroit, Michigan, February 24-27, 1997

Abstract: Results of the Charger Test Project conducted by Arizona Public Service (APS) indicated the importance of the charger/charge algorim/battery interface with respect to battery capacity. It also indicated that EVs rapidly charged demonstrated longer cycler life in field testing without reduction in capacity. Additional testing conducted by APS has shown that cycle life of the batteries is strongly related to the charging scheme utilized. Application of results of laboratory charging to the charge scheme utilized in this testing has been hampered by extremely long test periods in field tests designed to cycle batteries to the end of life. Development of a predictive method of cycle life extrapolation based upon microscopic analysis promises to reduce the time required for battery cycle life testing.

R.H. Newnham and W.G.A. Baldsing, New Operational Strategies for Gelled-electrolyte Lead/Acid Batteries, Journal of Power Sources 59, 137-141, 1996

Abstract: Gelled-electrolytes and flooded-electrolyte lead/acid batteries are cycled according to a new schedule that is based on partial-state-of-charge operation. With this strategy, the energy delivery of gelled-electrolyte batteries increases threefold, compared with that expected under traditional cycling procedures. In addition, overcharge is reduced by an order of magnitude. This provides charging efficiencies of up to 99.5%. Under the same duty, flooded-electrolyte batteries fail prematurely due to degradation of active-material at the bottom of the positive plates. This limitation is induced by increased active-material utilization as a result of acid stratification. The phenomenon is termed ‘localized PCL’, and is explained in terms of the uniform theory of PCL that has been developed by CSIRO.

R.H. Newnham, Advantages and Disadvantages of Valve-regulated, Lead/acid Batteries, Journal of Power Sources, 52, 149-153, 1994

Abstract: To improve the standard of living in many parts of the world, an efficient energy-storage medium with low-maintenance requirements is essential. Valve-regulated lead/acid batteries (VRBs) have the potentials to offer such a service at a competitive cost. Since there have been a few critical analyses of the efficacy of this technology, this discussion examines the advantages and disadvantages of using URBs in energy-storage applications. VRBs possess the following advantages: no water maintenance; little chance of acid spillage; negligible emission of acid and hydrogen; no special ventilation needs; minimal overcharge required at normal temperatures; easy to transport, and can be operated on their side. In addition, the gelled-electrolyte design of VRB is resistant to acid stratification, while the adsorptive glass-microfiber type has an excellent high-rate discharge performance. On the other hand, VRBs are not as resistant to overcharge as flooded-electrolyte units and must be charged very carefully to reach an acceptable service life. Other disadvantages include: specific gravity cannot be measured; a limited shelf life of two years, and an increased likelihood of thermal runaway.

D. Karner, R. Newnham, T. Gray, J. Wishart, N. Fengler, and X. Sun, PHEV Grid Interaction Evaluation. Assessment of Battery Technologies and System Configurations for Raps Duty.

September 1993

Two billion people are still without power to their homes. These people, who are remote from the mains grid, live mainly in the developing nations where abundant renewable energy from the sun and/or wind is available. Since the cost of grid connection can be prohibitive, the concept of the stand-alone, ‘remote-area power-supply system’ (the RAPS system) that incorporates a renewable energy source has been advanced. Following steady technological progress, RAPS systems are now becoming cost-effective in many situations. The design of such facilities, however, is far from optimum and still requires considerable development and refinement.The objective of this study is to evaluate the present status of RAPS battery and system technologies that will be eminently suitable for applications in developing countries. A subsequent target is the fabrication and field testing of those RAPS facilities that are identified as state-of-the-art, with a view to distributing proven technology in areas of need. The following aspects are reviewed: (i) individual system components; (ii) system configurations; (iii) laboratory and field studies of RAPS batteries/systems; (iv) manufacturers/distributors of such facilities; (v) availability and estimated cost of systems in developing countries. Special emphasis has been placed on both the battery and the charge-controller components, as these are considered to be the ‘weak links’ in present technology.Applications for RAPS systems have varied widely. Duties have ranged from wind-powered water pumping, back-up power units for microwave-repeater stations, to domestic electricity supplies. Typically, facilities comprise a battery bank (usually lead/acid), an energy source (solar, wind, micro-hydro or fossil-fuel-based) and a controller. The controller can vary in complexity from a simple mechanical switching device to a sophisticated microprocessor-controlled unit.The design of RAPS systems is determined by several factors. These include: (i) energy demand; (ii) availability of renewable energy; (iii) cost of system maintenance; (iv) site accessibility; (v) available finance. Either alternating current (a.c.) and/or direct current (d.c.) can be produced; the choice depends upon the quantity and quality of the power required. For sites with low energy requirements (i.e., 0.2 to 0.5 kWh/day), a d.c. photovoltaic (PV)/battery hybrid system is normally the preferred option. The same configuration is suitable for loads up to 2 kWh/day, but a small inverter is often included to provide a.c. power. Where more than 2 kWh/day is desired, PV/battery hybrid configurations are disadvantaged due to the high cost of the PV panels. At this level of energy consumption, diesel/battery hybrid systems become viable, provided that the operating costs of the diesel generator are reasonable. For this level of daily energy consumption, wind/battery hybrid systems are also competitive with solar-based designs, given that an acceptable wind regime exists (speed 6 m s-1).The study has involved the formulation of basic design specifications for charge controllers and RAPS batteries. This information will enable customers to acquire acceptable quality components. A detailed list of specifications for batteries has also been compiled to guide the user/consumer who requires ultra-performance from the energy-storage component.

A review of laboratory and field evaluations of different types of lead/acid batteries show that the performances of tubular- and flat-plate versions of both flooded- and gelled-electrolyte batteries, together with flat-plate absorptive glass-microfibre (AGM) designs, have been examined. The results reveal that the two flooded-electrolyte versions and the gelled-electrolyte type (assuming correct charging procedures) deliver acceptable service lives, both in the laboratory and in the field. By contrast, AGM units (at their current level of development) perform poorly in the laboratory and, therefore, are deemed not to be suitable for deep-cycle RAPS service where strict charge control is unavailable. The failure modes of batteries cycled in either the laboratory or the field are similar. Deterioration in performance is related mainly to degradation of the positive plate and, in some cases, to sulfation of the negative plates. Water loss from gelled-electrolyte batteries is negligible if correct charging procedures are adopted.

Manufacturers and suppliers have started to meet the needs of developing countries for RAPS systems. In general, present PV/battery hybrid systems for small houses cost between US$ 550 and US$ 900. These configurations, however, often employ automotive or modified-automotive batteries. It is common for such units to fail prematurely. Replacing this technology with imported, good quality, flooded-electrolyte or gelled-electrolyte RAPS batteries, could increase the system cost by up to $150 and $300, respectively. Nevertheless, the use of such batteries does not guarantee that the minimum accepted service-life of three years will be obtained, given the present level of sophistication and reliability of the control systems.

It is considered that the combination of an improved battery and a purpose-built charge controller, when introduced into small-scale PV/battery hybrid systems, would provide the durability required for the successful operation of such facilities in developing countries.

Analyses shows that calculations of the total energy costs associated with the operation of different RAPS system configurations should be interpreted with caution. Many assumptions and approximations are required for the modeling of RAPS systems, namely: (i) load profile; (ii) efficiency of energy generation device; (iii) inverter efficiency; (iv) battery bank efficiency; (v) fuel cost (diesel-based systems only); (vi) cost and lifetime of system/components; (vii) maintenance costs. Variations in these factors can have a marked affect on the overall energy costs.