Improved PV System Performance Using Vanadium Batteries

Robert L. Largent Design Assistance Division Centre for Photovoltaic Devices and Systems University of New South Wales, Kensington 2033, Australia

Maria Skyllas-Kazacos and John Chieng School of Chemistry and Industrial Engineering University of New South Wales, Kensington 2033, Australia

Abstract

A Vanadium-Vanadium Redox battery can improve Photovoltaic system performance, reliability and robustness by increasing the energy conversion efficiency of the battery to 87%, by making the battery life, efficiency and ongoing energy capacity independent of state of charge and load profiles and by reducing maintenance requirements. High battery efficiency reduces the required PV while a battery life insensitive to battery usage relaxes system constraints. These advantages are utilised in a demonstration PV system in Thailand that was designed specifically to use vanadium technology. Following a 12 month field testing programme with 4kW Vanadium Batteries, 300 systems consisting of 2-4 kW PV, a 4kW, 15kWh, Vanadium Battery and a 4 kVA grid interactive inverter are intended to be installed in residence in Thailand.

Introduction

in PV applications requiring energy storage, the selection of the energy storage system is of primary concern. The electrical parameters of the storage system constrain and shape the PV system. The deliverable power determines the maximum size of the electrical load, the energy storage capacity determines the duration of power to the load and energy conversion efficiency determines the amount of extra PV needed to make up the energy lost in the conversion. Additionally, the reliability of the storage system determines if the PV system can be used in a critical application and maintenance scheduling determines when personnel must visit the PV site.

PV systems engineers have traditionally employed electrochemical storage using lead-acid batteries. Extensive development and use of lead-acid technology, particularly in the automotive industry, has allowed the adaptation of that mature technology directly to PV applications. Lead acid technology is well understood, is reliable, is in mass production and is readily available; however, lead acid technology does have inherent attributes that must be designed around. In order to maintain energy capacity and long battery life, extra energy must be supplied periodically to the battery to de-stratify the electrolyte and to equalise the cell voltages. This process of "boost-charging" causes hydrogen evolution and water loss from the battery. The additional energy associated with this process is supplied by the installation of extra PV and periodic maintenance is used to replace the lost water. The battery life is strongly affected by how discharged the battery is allowed to get before it is recharged and, if the battery is allowed to stay in a discharged state for very long, irreversible damage occurs to the plates of the battery. A useful battery parameter, the state of charge, is difficult to determine accurately and after the battery is installed, it is, in practice, difficult to change the size of the battery to account for the addition of new loads not specified for in the original system design.

The constraints imposed by lead acid technology suggest that a more flexible, higher efficiency and cost effective technology would be a benefit to PV systems.

A new type of electro-chemical storage developed by the University of New South Wales (UNSW), the Vanadium-Vanadium Redox Battery [1], exhibits many of the qualities desired by PV systems designers. This battery has a very high efficiency, a reasonable electric density, high charge and discharge rates, a long lifespan independent of state of charge and load profiles, and low maintenance requirements.

These qualities greatly ease the constraints imposed upon PV system engineers. It is not neccessary to oversize the battery in order to maximise battery life or install additional PV for boost charging. During periods of low sunlight, the battery can be operated nominally at low states of charge with no effect upon battery life.

Additionally, this battery has a feature which allows for many new options not available with lead acid technology. It is possible to simultaneously charge the battery at one voltage while discharging it at another voltage. This feature can be utilised to make a minimum cost, high efficiency, maximum power point tracker or allows to operate the battery as a DC transformer, electro-chemically transforming a current and a voltage into a different current and voltage.

The Centre for Photovoltaic Devices and Systems in collaboration with the UNSW Vanadium Research Group and the Thai Gypsum Products Co. Ltd., Thailand, has designed and installed a PV system using Vanadium Battery storage in a demonstration house in Thailand. This is a pre-commercial prototype version of a residential grid interactive system intended for installation in 300 houses in Thailand.

The Vanadium Battery

Redox flow batteries employ a different energy conversion method than solid plate batteries. In contrast to the solid phase chemical changes that occur on the plates of a lead acid battery, a redox battery stores energy as chemical changes in two liquid electrolytes that are hydraulically pumped through the battery stack. Energy conversion occurs in the battery stack and the charged electrolyte is stored in reservoirs external to the battery stack. The physical size of the battery stack determines the power available from the battery and the volume of the electrolyte reservoirs determines the kWhrs energy storage of the battery.

picture:vanadi1.gif

[IMAGE]

Figure 1, Schematic View of Vanadium Cell

The development of the Vanadium-Vanadium Redox Battery at UNSW has overcome significant technical limitations that have plagued other type of redox flow batteries. [2] The battery is insensitive to atmospheric oxygen, has a high 1.4 V cell voltage, has high longevity, low maintenance requirements and the electrolytes are not mutually destructive.

In the vanadium battery, identical electrolyte is used initially in both the positive and negative sides of the battery. During charging, electro-chemical reactions within the battery stack change the valance of the vanadium in the two electrolytes with the negative reaction changing V(III) to V(II) and the positive reaction changing V(IV) to V(V). This process is reversed during discharge. If any inadvertent mxing of the charged electrolytes occurs there is an energy loss as heat but, because the mixed electrolytes revert back to their uncharged states, they can be recharged next time through the stack. Thus, cross contamination is not detrimental to the longevity of the battery. The above reactions do not, under all normal operating conditions, generate hydrogen.

Because both the valance reactions are permissible to the originial electrolyte, it is an arbitrary decision as to which side of the battery is positive and which side of the battery is negative. Only after initial charging is there a positive and a negative side of the battery.

The battery stack's electro-chemical reactions are all highly efficient with the energy voltage and colombic efficiencies ranging from 90 to 99%. When the energy needed to operate the pumps, which amount to 2 to 3% of the total battery energy, is also taken into account the total battery efficiency is a very high 87%.

An accurate state of charge determination is made possible by measuring the open circuit voltage of a small vanadium cell attached to the battery with some portion of the electrolytes being pumped through it.

These attributes of the Vanadium-Vanadium Redox Battery make its use in PV systems very desirable.

PV Systems with Vanadium Batteries

The use of a Vanadium Battery with its very high energy conversion efficiency and no boost charge requirements directly relates to less PV being needed for the system. Greater system robistness is achieved through the battery's ability to be left indefinitely at any state of charge with no reduction in battery life and, because there is no hydrogen evolution, there is no water loss from the electrolytes. Greater system flexibility is achieved with the new capability of tailoring the kWhrs storage to meet any new loads by varying the volume of the electrolytes and, because the electrolytes are stored separate from each other, there is very low self discharge. The battery itself can supply multiple output voltages -- a valuable advantage in PV systems with DC loads of different voltage requirements. These features offer new versatility in the choice of applications that use PV systems. Material redundancy is minimised by the full power, very deep cycle (100%) capability of the battery and additionally, because there is no hydrogen evolution, there is no need for forced ventilation. Maintenance requirements are low which reduces visits to PV installations.

 

Pump Losses

The 2-3% energy loss associated with the vanadium battery's pumps is calculated with the battery operating at full power. If the battery is operating at low power then the pumping power loss is a more significant proportion pf the system power.

The strategy to minimise this energy loss and improve system efficiency is to turn the pumps off during periods of low charge or discharge rates. With no electrolyte flow all of the power going into or coming from the battery operates directly on the electrolytes present in the stack.

When the energy level of the stack eletrolytes reaches a threshold, the pumps are turned on for a period of time which fills the stack with fresh electrolytes, then the pumps are turned off and the battery again waits until the treshold it met.

In PV systems where random load profiles are present this feature allows for an optimisation of pumping energy versus system load power requirements. UNSW is developing a micro-controller based vanadium battery controller with strategies for optimising the efficiency of the battery for these applications.

The ability to easily and accurately determine the true state of charge of the Vanadium Battery allows for dynamic predictions of the amount of time that a battery can sustain a load. This allows greater system diversity and gives the designer the ability to fine tune the kWhrs storage of the battery for differing load profiles and load types.

 

Voltage Taps

A valuable feature of the Vanadium battery is its ability to have its charge voltage being different than its discharge voltage. It is possible to simulateously charge the battery at the 12V tap and discharge at the 48V tap or visa versa. Used in this manner, the battery becomes an 87% efficient DC transformer.

picture:vanadi2.gif

[IMAGE]

Figure 2: Voltage taps for MPPT and differing charge and discharge voltages.

Voltage taps increase system flexibility as loads with different DC voltage requirements may be operated from one power source without the additional conversion losses associated with voltage matching.

 

Maximum Power Point Tracking

A maximum power point tracker (MPPT) is useful in reducing the PV required in system applications. The relatively high cost of the power electronics MPPT, however, often reduces the cost effectiveness of the reduction of PV.

The tap change method presents itself as being a highly cost effective and efficient method of Maximum Power Point Tracking.

PV array's maximum power point can be matched buring charging by choosing an appropriate voltage tap on the Vanadium Battery and changing to another voltage tap as PV array's maximum power point changes. Unlike the complex power electronics counterpart, there are no energy conversion losses associated with the tap change method and the electronic are relatively simple and rugged.

Economic Considerations of the Vanadium Battery

An economic analysis of Vanadium Storage technology has determined the cost of Vanadium electrolytes to be US$48/kWhr and the cost of the battery stack components to be US$206/kW. Using a factor of 2.5 to account for the additional costs of storage tanks etc., resulted in the capital cost for a battery varyinig from US$635/kWHr for a battery with 1 hour storage capacity at full power discharge (e.g. 4 kW battery with 4 kWhrs storage) to US$146/kWhr for a battery with 20 hours of storage capacity at full power discharge (e.g. 4 kW battery with 80 kWhrs storage). This cost analysis indicates that the cost per kWhr is determined by the ratio of the battery's power output to the number of total hours of full power storage. Thus both a 1 kW battery with 20 kWhrs storage and a 4 MW battery with 80 MWhrs storage would be US$146/kWhr.

picture:vanadi3.gif

Total Cost of Vanadium Storage per kWhr as a function of Storage Time

[IMAGE]

 

Figure 3: Capital Cost of Vanadium Battery per kWhr

Most obvious in this analysis is the dramatic drop in cost per kWhr as the battery goes from 1 to 5 hours of full power storage.

A major economic advantage that Vanadium technology has over other technologies relates to the ongoing costs of battery storage. Because the electrolytes are not damaged by atmospheric oxygen or cross contamination they have an indefinite life and are considerd to be a capital cost. Current estiomates indicate that the battery stack will need to be replaced every five years yielding a very low ongoing cost, when contrasted with a lead-acid battery where, in PV applications, the entire battery needs to be replaced on the average of every seven years.

PV & Vanadium Demonstration System in Thailand

The first licensee for the commercialisation of the Vanadium Battery is the Thai Gypsum Products Co. Ltd., (TGP) Bangkok, Thailand. TGP built a PV & vanadium demonstration house on their industrial estate at Laem Chabang, Thailand with the opening, lead by HRM Princess Maha Chacri Sirindhorn of Thailand, on 23 December 1992. This function has 600 guests from industry, military and the media.

This PV & Vanadium system was installed in the demonstration house in December 1992 by members of the UNSW Centre for Photovoltaic Devices and Systems, UNSW Vanadium research Group, and the Thai Gypsum Vanadium Commercialisation Group.

The demonstration system was designed to operate AC loads and a small, less than 800 watt "compressor type" air conditioner was chosen as the load in the demonstration house.

picture:vanadi4.gif

[IMAGE]

Figure 4: Schematic of PV and Vanadium Battery system installed in Thailand

 

Construction and assembly of the system was as follows:

In Thailand, TGP built the demonstration house and roof mounted 36 Kyocera LA441K63 PV modules giving 2.2kW of installed PV.

In Australia, the UNSW Centre for Photovoltaic Devices and Systems selected the inverter and other system components and built the micro-controller for the Vanadium Battery as specified by the UNSW Vanadium research Group.

The UNSW Vanadium Research Group designer and built a Vanadium battery rated at 1.2kW, 15kWhrs. This battery has 12 cells, giving a system voltage of 16.8 Volt, and uses 200 litres in each of the two lectrolyte reservoirs.

Butler Solar Products, Australia, the designers of the Siemens' range of SUNSINE inverters, modified an existing 1 kW, 12 Volt stand alone SUNSINE inverter for the 16.8 Volt PV & Vanadium system. This required a redesign of the transformer, installation of additional FET's in the bridge arms and modifications for the Thai requirements of 220V, 50 Hz output. This inverter is not grid interactive.

The system load is a National CU-700K split system "compressor type" air conditioner. The starting power required for this air conditioner was measured to be from 6-11 kW -- a considerable amount of peak power for a 1 kW system.

The initial charging of the Vanadium Battery was with a power supply connected to the AC grid.

The demonstration system works as designed. Work continues with this system giving the Thai Gypsum Vanadium Commercialisationi Group hands on systems experience that will be directly applicable to their 300 house project.

 

300 House Project

TGP is in the process of commercialising thr 4 kW vanadium battery with the first application of the technology being a 300 house installation in Bagkok. Each house will have a PV & Vanadium Battery system consisting of 2-4 kW PV, 4kW Vanadium Battery and a 4 quadrant, 4kVA grid interactive inverter. It is hoped that the first of the houses will be complete by the end of 1993 and that all 300 will be complete 18 months later. Data acquisition for system evaluation will be employed.

Conclusion

The Vanadium-Vanadium Redoc Battery offers system performance benefits through increased system efficiency and robustness, reduced maintenance requirements, and greater flexibility in both system design and system application.

The 300 house residential grid connected systems will test this Vanadium technology in a variety of system configurations.

 

Achknowledgements

The Centre for Photovoltaic Devices and Systems is supported by the Australian Research Council under the Special research Centre Scheme and by Pacific Power.

Research for the Vanadium Battery development has been funded by ERDC, NSW Office of Energy and Thai Gypsum product Co., Ltd. the support of Formica Australia is also gratefully acknowleged as is the assistance of Michael Kazacos, Rui Hong, Dennis Yan and Jim Wilson.

References

[1] M.Skyllas-Kazacos and R.G.Robins, "All Vanadium Redox Battery", US Patent No 4 786 567, 1988

[2] Maria Skyllas-Kazacos, D. Kasherman, D.R. Hong, and M. Kazacos, "Characteristics and Performance of a 1 kW UNSW Vanadium Redox Battery", journal of Power Sources, vol. 35, 1991. pp 399-404