Vehicle to Grid (V2G) Energy Transfer
A seductive idea, but what's in it for me - (or anybody)?
The V2G concept currently under serious consideration as a Smart Grid application is to use suitably equipped private, passenger electric vehicles to provide a load levelling function, or temporary energy storage, for the electricity generating utilities. The idea is to charge the vehicle's batteries at night when demand is low utilising the utility's excess generating capacity, so called "valley filling" of the utility's load profile and allowing the utility to draw energy back from the battery into the grid as required during the day when demand is high to reduce the peak demand on the utility's generating capacity, known as "peak shaving". It may sound very attractive in theory, particularly for the utilities, but a lot of things need to happen for it to be put into practice.
Quite apart from the inconvenience to the vehicle owner, it would require a network of public charging stations capable of bi-directional power transfer, each station incorporating an inverter with precisely controlled voltage and frequency output to feed the energy back into the grid. It would also require the support of a massive communications network to manage the distributed power flows, the billing and feed-in buy back transactions.
It is expected that users will be tempted by advantageous pricing and electricity buy back tariffs and in return, to allow the use some of the available cycle life of their expensive batteries for temporary energy storage to service the grid for the benefit of the utilities. How much sooner will you have to replace your battery? What happens if your regular work schedule is suddenly disrupted and you return to your car two hours earlier than usual, only to find that the utility has drained all the energy from your battery? How will the resale value of the vehicle be affected?
Up to now there has been no discernible consumer demand for this service.
V2G looks great till you start to put numbers on it.
Let's take a typical example to look at the facts of this "opportunity' and the alternative of simply providing increased generating capacity.
What's in it for the EV owner?
First the economics: There should be some benefit to the EV owner in compensation for the use of his battery and the possible inconvenience which that incurs.
The Energy Cost
A typical EV battery has a capacity of around 25 kWh and a lifetime of 2000 cycles and costs about $10,000. Although it has a capacity of 25 kWh, its depth of discharge is usually limited to around 80% to improve its life. This means that when it is fully charged, it has only 20 kWh of available energy. The cost of charging the battery is not just the cost of the 20 kWh of energy in the battery since there is typically an efficiency loss of 10% in the charger used to convert the AC from the generating company to DC used to charge the battery. Thus the user has to purchase 22 kWh of energy to provide the 20 kWh of available energy. With electricity costing ten cents per kWh at the point of use, it costs only $2.20 to charge the battery for a day's commuting, and this is the revenue received by the electricity generating utility. But this is not the only cost to the user. There’s also the share of the capital cost of the battery.
The Battery Depreciation
Each charge - discharge cycle of the battery consumes one of the battery's 2,000 cycles or one two thousandth of its lifetime. With the battery costing $10,000, this amounts to $5.00 per cycle and this is much more than the cost of the energy stored in the battery.
Total Cost of Energy Supplied
Taking depreciation into account, the total cost of charging the battery with 20 kWh of available energy is $7.20 which amounts to 36 cents per kWh.
V2G Energy Available to the Electricity Utility
If the EV owner wants to sell the 20 kWh of stored energy back to the utility, it must first be converted back to AC in an inverter. Just as in the charger, the inverter typically has an efficiency loss of about 10% so that the amount of energy available to the utility at the point of connection is only 18 kWh.
Total Cost of V2G Energy Available
Taking the inverter efficiency into account, the user has only 18 kWh of available energy which cost $7.20, so the equivalent cost of the energy available to sell back to the generating company is 40 cents per kWh. This is the breakeven price or the minimum feed in tariff needed by the EV user just to avoid a loss.
Summary of the V2G Proposition for the EV User
Users must accept that every charge-discharge cycle used by the utility will mean one less available to themselves for travelling and the available cycles will be shared with the utility on a pro rata basis to their use. Since the utility usually will need the V2G energy during the day when demand peaks occur, not during the night, commuters must arrive at work with sufficient charge left in their batteries to make this energy exchange feasible. Otherwise they must specify oversize batteries which are already very expensive.
It is possible to program the Smart Grid or V2G system to ensure that there will always be sufficient energy available in the battery to get home at the end of the working day, or even to prevent discharge of the battery if it is planned that the car will be needed during the day, but this does not help if the user is called away unexpectedly during working hours to attend to some domestic or commercial emergency and the utility has just drained the battery.
Because the battery is usually specified to last the same length of time as the vehicle when used under normal driving conditions, using the available cycles to satisfy the demand of the grid will result in fewer cycles being left to power the vehicle. This will result in the battery needing to be replaced before the vehicle has reached the end of its life. Who would want to purchase a $10,000 battery to put in a used car which has only one or two years of useful life left?
Consumer Economic Summary
The consumer buys 22 kWh of energy at 10 cents per kWh but only has 18 kWh to sell back to the utility and must obtain a price of 40 cents per kWh just to break even. To be attractive to the consumer, and to compensate for the inconvenience, an economic selling price would be more like 55 to 60 cents per kWh or more. Is it likely that the generating company will pay such a price when it doesn't cost them anywhere near that to produce the energy in the first place?
It looks an unlikely proposition, but maybe the cost to the utility of satisfying the peak power demands by alternative means can justify this high price. We need to look at the utility's economic model.
What's in it for the Generating Company?
The generating company's objective is to find the most economical way of satisfying the varying consumer demands for energy which are characterised by very high peak demand during the day and a much lower demand during the night. They have two options: Install enough capacity to supply the peak daily demand which results in the equipment being under-occupied during the night resulting in poor utilisation of their assets. Or they can try to even out the demand by shifting some of the daytime peak load to the night time so that it can be generated when the equipment is only lightly loaded thus achieving much better asset utilisation.
The following examples explore the two alternatives available to a local generating company operating a 500 MW gas fired generator, with a capacity factor of 85%, generating 3723 GWh of electrical energy per year (8760 hours). This is enough to supply a typical (USA) surrounding community of over 350,000 households, with each household consuming around 10,000 kWh per year. (More if they charge their EVs from their domestic supply.) The utility's base line annual sales revenue at 10 cents per kWh of energy generated will amount to $372.3 million, with corresponding unit generating costs of 5 cents per kWh (Source IEA. See below), giving rise to total generating costs of $186.2 million and resulting in a gross margin of $186.2 million.
Problem - Inability to meet peak demand requirements
We can start with an arbitrary example of a resourcing problem the utility might have to solve which is:
How can they supply a peak demand, which is 5% more than their installed capacity, for 1% of the time? This equates to a power shortfall of 25 MW for 87.6 hours or 25 MW for 1 hour, 87 times per year or a total energy shortfall of 1862 MWh for the whole year.
Solution 1 - Install a new gas fired generator
The typical cost of a gas fired generator is $500 /kWe (Source IEA - $400 to $800/kWe), so a 25 MW generator of the same type as those already in use would cost $12.5 million. This would provide the necessary 5% increase in capacity and at the same time would increase the value of the installed capital equipment by the same 5%. The plant utilisation will however be reduced by 5% from 85% to 80%.
So how will the new capacity affect the generating costs?
For the purposes of the discussion we assume that both the "original" plant and the incremental "new" plant will both depreciate at the same rate.
Operating cost assumptions from IEA:
The capital cost as percentage of total generating costs is less than 15% and this is mostly due to depreciation. The Operations and Maintenance (O&M) costs are less than 10% of the total generating costs. Fuel costs are around 80% of total generating costs and the typical generating cost of the electricity produced is 5 cents per kWh. This is the cost to the utility at the point of generation. There will however be resistive losses of typically 7% in the distribution network before the energy reaches the consumer, so that the cost to the utility of supplying 1 kWh of energy at the point of use, where it is measured and billed, will be 5.35 cents.
The fuel cost per kWh of the incremental electrical energy produced will be the same as for the base load since there should be no change fuel efficiency if the generating equipment is the same.
The increase of 5% in capital investment will result in an increased depreciation charge and this will apply to all the energy generated by the plant and since the capital costs typically contribute to just 15% of total generating costs, this would result in an increase of 0.75% in generating costs. This amounts to an increase of 0.0375 cents on the 5 cents operating costs.
Besides this the incremental plant will also incur additional Operations and Maintenance costs. As above, this will increase the 10% of the generating costs attributable to O&M by 5% thus increasing the generating costs by a further 0.5% or 0.025 cents.
Thus the total generating costs per kWh will increase from 5 cents to 5.0625 cents due to the cost of purchasing and operating the extra capacity needed to supply the peak loads.
Summary Solution 1 New Plant
For a capital cost of $12.5 million, the utility will be able to supply the extra 1862 MWh of energy needed during peak demand periods. This will in turn generate incremental revenue of $186,200 at 10 cents per kWh. The total energy generated will increase to 3724.9 GWh and with a generating cost of 5.0625 cents per kWh this will cost $188.6 million. This compares with $186.2 million for running the plant without the increased capacity, an increase of $2.4 million in costs to generate an increase in revenue of only $186 thousand. Because it will only be needed during peak periods the incremental plant will be idle for more than 99% of the time. So far it looks like a poor deal for the utility.
Solution 2 - Provide Vehicle to Grid Infrastructure
Even before we consider the infrastructure needed to implement Vehicle to Grid energy transfer, we need to consider suitability of the vehicles to store and return the required energy. The ideal vehicle is a pure EV with a typical battery capacity of 25 kWh or more. As noted above, to avoid over-discharging, they could make 20 kWh of DC energy available to the utility, but taking into account inverter losses of 10% and a further 7% distribution losses to get the energy back to where it is needed, the actual energy available to the generating company at the new point of use will be only 16.6 kWh. With generating typical costs of only 5 cents per kWh, the baseline value of the total energy fed back to the generating company will be only 83 cents. Peak energy is obviously worth much more than this, but it at least the base line gives us a reference point.
One further consideration is that in reality, well over half of the vehicles participating in the V2G programme will be PHEVs with battery capacities typically less than 15 kWh. This means that the net available stored energy after taking losses into account will be less than 10 kWh so that almost double the number of vehicles will be required to provide the necessary storage capacity and the base line value of the total energy available per vehicle will be less than 50 cents. The situation with HEVs is even less attractive. They typically have batteries with less than 2kWh capacity of which after losses only half would be available to the generating company. The base line value of the total available energy stored in an HEV battery will be around 5 cents which makes V2G energy supply impractical or uneconomic for HEVs.
The huge cost of providing the infrastructure necessary to harvest these very small, low value, dispersed amounts of energy needs to be seen in this context.
The purpose of the V2G infrastructure is to be able to call on available energy sources in response to demand peaks. To do this the utility needs to be able to discharge the EV batteries as fast as possible to follow the demand profile. For this reason the capacity and installed cost of a V2G feed-in station will be similar to a level 3 charger at $20,000 minimum.(Assume only $10,000 in this case)
In the example above, with 16.6 kWh of energy available from each vehicle, it would need 1506 electric vehicles to deliver 25 MW for one hour, or double the number of PHEVs, and these would all need to be within the operating region of the generating company. But most vehicles will not be fully charged when the morning peak demand occurs having just made the journey to work and many feed-in stations will not necessarily be occupied when the energy is needed. If the demand peaks occur during the morning or evening rush hours, most vehicles will be travelling and not available to feed energy back into the grid.
Alternatively, on a power basis rather than an energy basis, to provide the maximum instantaneous power with the minimum number of vehicles or batteries, the batteries must be discharged at the maximum C rate specified by the battery manufacturer and this is controlled and limited by the vehicle's on board battery management system. A 25 kWh battery could possibly deliver a power of 50 to 100 kW depending on the manufacturer. This would fully discharge the battery in 15 to 30 minutes. To deliver the instantaneous peak power of 25 MW needed in the above example with 25 kWh batteries with a typical power output of 75 kW would need a minimum of 333 batteries, ignoring losses. If the duration of the power demand exceeded about 20 minutes then proportionately more batteries would be needed. This is not an ideal solution since discharging batteries continuously at their maximum C rate has an adverse effect on their life. Discharging at the more benign 1C rate or below will also need proportionately more batteries to deliver the peak power.
The net result is that a network of at least 5000 charge-discharge feed-in stations (assumption, possibly many more if vehicles are primarily PHEVs) will need to be constructed at work places, car parks and private residences to provide the necessary V2G capacity. Since it is expected that most charging will occur during the night, most users will need to install intelligent charging stations at home where overnight charging normally takes place rather than depending on public feed-in stations. Furthermore, because of the low market penetration of electric vehicles, and the random availability of vehicles at feed-in stations when and where they are needed, almost every EV and PHEV owner would have to sign up to the V2G programme to fulfil the necessary peak energy requirement.
Assuming just 5000 feed in stations and 1500 vehicles needed to satisfy the peak demand, the occupancy or utilisation of the feed in stations during the peak periods will only be 30% and since these peaks only occur 1% of the time the overall occupancy will be a miserable 0.3%. Taking into account the charger and inverter efficiencies of 10% each and the 7% each way distribution losses in the network, the round trip efficiency of the temporary energy storage in electric vehicles is only 70%, a 30% energy loss and a truly wasteful use of fuel by the generating company. Providing emergency generating capacity or "valley filling" by means of V2G energy transfer thus increases the generator's fuel costs by 30% and since fuel costs are 80% of generating costs this equates to 24 % of total generating cost or 1.2 cents on 5 cents resulting in a cost of 6.2 cents per kWh of the incremental energy provided by V2G.
But that's not all. There's a massive capital cost needed to upgrade the electricity distribution grid to provide the V2G infrastructure. This includes replacing existing charging stations with bi-directional feed-in stations, upgrading the network infrastructure with inverters in each feed-in station, providing power switching devices to control the bi-directional energy flow and upgrading safety systems to avoid hazards to grid employees working on the network when users connect their batteries to the network. In addition an intelligent computer network will be needed to follow agreed customer load profiles, to monitor the charge in the batteries, to manage the energy flows and to ensure that the vehicles always have enough energy to get home. An authentication and billing system will also need to be set up. These requirements will result in increased operations and maintenance costs to cover the cost of managing the new infrastructure.
On the positive side, any incremental "valley filling" opportunities will improve the utility's plant utilisation. Transferring the excess demand of 25 MW for 87.6 hours to a period of low demand increases the annual plant output by 2.19 GWh. With a base load of 3,504 GWh per year, this amounts to an improvement in plant utilisation of 0.72%
Summary Solution 2 V2G
There is still a massive gap between the cost of generating the energy and an economic feed in price of 55 to 60 cents per kWh needed to make this an attractive proposition for the consumer. This is made even worse by the costs of installing and operating the necessary infrastructure to make it happen. The current energy generating cost per kWh of 5 cents will increase to 6.2 cents if the energy is supplied via V2G due to the extra fuel needed to compensate for the round trip losses of the V2G system. In addition the total new infrastructure costs to provide 5000 V2G feed-in stations at $10,000 each will be at least $50 million. This alone is four times the cost of providing the alternative, extra generating capacity and will increase the capital cost of supplying all the energy by 3% and the cost per kWh of all the energy supplied by 0.15 cents.
Despite the lack of commercial justification, academics, consultants, systems management, computer and automation companies continue to promote the introduction of V2G systems possibly with the hope of winning contracts to carry out feasibility studies and systems development in pursuit of tempting energy efficiency improvements.
There are some circumstances in which V2G applications could reasonably be justified. A military operation in theatre usually needs a totally independent and self-sufficient energy supply, a so called “Island of Power”. It needs to provide power for mission critical applications but does not necessarily have the luxury of generating capacity to accommodate all power peaks. In such an application where security of supply is much more important than costs, a V2G system could be used to provide the emergency power capability. The troops however would not be too pleased if they suddenly come under attack, only to find that the batteries in their vehicles had just been drained to provide the power for cooking their lunch.
Buffer Storage for Wind Energy
This is the continuing problem of providing a constant, dependable energy supply from an intermittent energy source. Up to now this is being solved by installing reserve generating capacity to supply the load when the wind is not blowing. V2G energy transfer has been proposed as a solution to this problem but this suffers from the same problems as V2G being used for simple load levelling as outlined above. Furthermore, the system has to cope not just with variable demand but also the uncontrolled variability of the supply. Thus the battery in the vehicle may be called upon, under the control of the generating company, to store excess energy generated from the wind during the night when the demand on the grid is low. This becomes a much more complex energy management problem and there is still little or no obvious incentive for the EV owner to accept the inconvenience just to help the power utility to solve its energy management problems.
It would appear that while the idea of V2G energy transfer would provide a technically feasible load levelling function, there is no economic incentive to either the EV owner or the electricity generating utility to provide such a system. The value to the utility company of the total energy available from each electric vehicle is less than one dollar and less than 50 cents from PHEVs. Can this be a serious proposition?! A massive network of expensive feed in stations would need to be installed to harvest this energy and thousands of users would need to sign up to satisfy the expected peak. To be attractive to the consumer, the feed in price per kWh of the energy delivered back to the utility is 55 to 60 cents or over twenty times the utility's baseline cost 5 cents for generating that energy. Unfortunately there doesn't appear to be a way of bridging that gap.
These conclusions have been drawn from an arbitrary peak energy shortfall of 5% for 1% of the time and they show an overwhelming economic case against the introduction of V2G energy transfer. To accommodate greater or more prolonged peak energy demands would require an even larger number of participating users and a correspondingly larger feed in network infrastructure, making the case against V2G even stronger.
Though not all the costs have been quantified, some gross assumptions made and some minor effects have been ignored, even if the costs of installing increased capacity have been underestimated by a factor of ten, there would still be no economic justification for V2G energy transfer. It is quite clear that installing more generating capacity is by far the simplest and least expensive solution to dealing with peak demand, though the downside of this solution will be a slightly reduced capacity utilisation.
As is usually the case - The simplest ideas are the best.
So it should not come as a surprise that there's no queue of takers for this wonderful opportunity.