It’s been a while since the last post in this series on electric vehicles (here are parts one, two and three), but this post is number four. Today, I’m looking at the costs of these cars – both their running costs, and their capital costs. Again, I’ll abbreviate plug-in hybrid electric vehicles to PHEVs, and battery electric vehicles to BEVs – these are the “full” electric vehicles which don’t have an engine for backup.

This post is about the cost of electric vehicles – the main reason they’ve been so slow to take off. These cars are much more expensive than conventional cars, unless there are hefty subsidies involved.

Capital (“Up Front”) Costs

The high capital cost of EVs is driven in large part by the batteries. The latest generation of vehicles use lithium-ion batteries, which are much better at storing energy than the traditional lead-acid batteries you’ll find in your Corolla. They’re also much more expensive, although the price is falling and will continue to do so. The graph below shows some scenarios for price decline:

BEV battery cost forecasts

Battery costs are usually measured in terms of a cost per kilowatt-hour (kWh) of energy storage; a PHEV might have a battery with 8 kWh, and a BEV might have 30 or 40 kWh. When I was writing my thesis a couple of years ago, costs of up to USD $1,000/kWh were being floated around, although there was and continues to be a wide range of different opinions. Adding to the uncertainty, early EVs will have been sold below cost, or at least at less-than-economic returns to the manufacturer, as they started to develop the technology. It seems to be generally agreed that battery costs are now less than USD $500 per kWh, although manufacturers would obviously want to make a profit on those costs at some point, and there are taxes and other considerations as well.

So, what kind of price difference would that mean for a new PHEV or BEV in New Zealand? Let’s say that the car manufacturers are happy with a battery selling price of USD $500 per kWh, around $570 in NZ dollars. Adding GST onto that brings the figure to around $650. Therefore, an 8 kWh PHEV battery could cost $5,200, and a 33 kWh BEV battery might be around $21,450 – still not cheap by any measure. Things get a little less straightforward when you consider that the PHEV will cost a little more due to having both an electric motor and an engine, and the BEV will cost a bit less since its electric motor is quite a bit cheaper than the typical engine.

Running Costs

As discussed in part two, electric motors use a lot less energy than a traditional car engine. This means lower running costs. But how much lower? From my earlier posts, a vehicle running on electricity could use around 20 kWh to travel 100 km. To see how much that costs, simply look at your power bill. Across New Zealand, households pay an average of 28 cents per kWh, according to the MBIE. The “marginal” cost you’ll pay for an extra unit of electricity, though, will be a bit lower. I’ll use a figure of 22 cents per kWh.

This gives a cost of $5 per 100 km – certainly much cheaper than a typical petrol car, which uses 10 litres of petrol to travel 100 km, costing around $22.00 at current petrol prices.

However, a big chunk of the petrol price is tax, comprising a contribution to the National Land Transport Fund, and a bit to ACC as well. According to the MBIE, that’s around 77 cents per litre once GST is added on, or $7.70 per 100 km. Since EVs also contribute to road wear and tear (and demand for new investment), and to accidents, they should also be paying something for this. We obviously can’t tax them through petrol, and it’d be pretty hard to do it through electricity prices as well, so the logical way to do it is through Road User Charges. Indeed, EVs would normally be subject to these, but they’ve received an exemption for the time being (to encourage their uptake). Perhaps that’s a sensible move, but it’s probably not something we’d still want to do in 20 years time when a growing number of cars are electric, and drivers of old cars will need to pick up the slack and pay more tax.

As I’ve written previously, the long-term solution may be to make Road User Charges universal, although there are issues with this as well. For now, I’ll just note that EVs might either be exempt from Road User Charges (i.e. not directly contributing to the upkeep of the transport network and accident costs), or they might end up paying the full charge. This would more than double the running costs of BEVs, although they’ll still be cheaper than petrol cars.

Sitting awkwardly in the middle of all this are PHEVs. At the moment, they get a somewhat inconsistent treatment. Petrol-electric hybrids, for the time being, pay tax through their petrol consumption. In my thesis, I assumed they average 3 litres of petrol per 100 km, although this will vary substantially. Drivers who only do short trips could end up using the electric motor for nearly all their driving. Regardless of the actual figure, they may end up paying very little tax.

Diesel-electric hybrids, on the other hand, have to pay Road User Charges, so they end up paying the full whammy of costs (once the RUC-petrol tax discrepancy gets resolved in the next few years). That’s a real disincentive from buying diesel-electric PHEVs, so we’d expect them to be much less popular here.

The graph below compares the lifetime running costs of several kinds of car, under several taxation scenarios. As you can see, RUCs or the lack of them make a big difference. The Excel file is here if you want to play around with it.

Lifetime opex

Getting the costs to stack up

Setting aside environmental concerns, “range anxiety”, and all the rest, consumers will be prepared to pay the higher capital cost of electric cars, if they’re going to save enough money on their running costs. In the graph here, for a car travelling 12,000 km a year for 25 years (perhaps a bit on the high side), and using an 8% discount rate, you’ll pay nearly $30,000 in running costs for a petrol car, compared with $7,000 for a BEV which is exempt from Road User Charges forever.

That’s a $23,000 difference, for quite an extreme case. For some of the other BEV/ PHEV combinations, the difference is $10,000 to $15,000. The difference would get smaller with a higher discount rate, or with less travel.

Overall, if you compare these running cost savings to the extra capital cost, it looks like the financial argument for BEVs and PHEVs isn’t quite there yet.

Wrapping Up

Battery costs will continue to decline, driven by economies of scale (i.e. production scaling up) and technological advances. It’s hard to predict how fast costs will come down, or by how much. Someone might invent a transformational new battery chemistry (rather than lithium-ion), or we might simply see incremental advances.

There are ways of reducing this issue: for example, customers could lease electric vehicles, or buy the vehicles but only lease the batteries. This kind of scheme could allow the buyer to avoid the high up-front cost, which could be recouped over time through the running cost savings. Electricity providers would find this a straightforward extension to their business, and I believe a number of companies in New Zealand would look at running these schemes.

At current price levels, BEVs have running costs that are only marginally lower than petrol-electric PHEVs, because these hybrids are only taxed on their petrol consumption. Furthermore, even though diesel-electric PHEVs will be more efficient than petrol-electric PHEVs, they are likely to have higher running costs.

BEVs currently have an exemption from Road User Charges, to encourage their uptake over the next few years, but there’s no reason why this should be the case in the long term – they use the road network, and should pay their share.

Since the costs associated with the road network are primarily dependent on the weight and number of vehicles using the road – and not on the litres of fuel used – the Road User Charges scheme arguably provides a more equitable way of charging for road use.

Share this


  1. Wouldn’t the annual opex for cars increase as they age due to the need for ongoing repairs etc, rather than decrease as the graph suggests?

    1. [This is partly a response to Greg below as well] Sorry, perhaps I should have been clearer… I’m only talking about the costs of petrol/ electricity/ RUCs in the graph above. Not including depreciation, maintenance, tyres etc. Electric motors are pretty low maintenance, so for a BEV you’d save quite a bit on maintenance… until you have to replace the battery, whenever that might be. Of course, when it’s time to replace the battery (maybe 10 years down the track, with continued manufacturing cost improvements) you’ll hopefully be paying a bit less for it.
      There’s an argument that EVs might depreciate slower than conventional cars, excluding the battery (which you replace anyway), since there are fewer other parts of the car that are getting run down. On the other hand, maybe later generation EVs will be so much preferred to today’s ones that they’ll actually depreciate quickly.
      It’s hard to quantify these issues so I’ve left them out. Also note that I’m using NPVs in the running costs graph, which is why the curve flattens out over time – you’re using the same amount of fuel, driving the same distance, but it’s being more heavily discounted.

  2. The big question with any Battery EV is how long will the battery pack last 10 years? 15 years at best and what will it cost then to replace?

    As if it needs replacing even once in its lifetime, it totally changes to economics of BEVs versus the others (making it even more uneconomic).

    Right now BEVs don’t stack up financially because they are too simply expensive due to the costs of the batteries and thta assumes that the battery never needs replacing.
    Problem is that Petrol/Diesel are actually great energy storage device – whether per Kg or “Litre”, so any replacement technology has to overcome a lot of physics and chemistry issues to make petrol or diesel fuels obsolete for powering vehicles.

    Of course, if for instance we had wireless energy transmission in the roadway so that for example BEVs could have small batteries that are semi-continuously charged from from the grid as they drive on the roads, that would change the economics in their favour a lot. But that would require a lot of investment to achieve.

    Then of course, there are also similar technology for trams and trains (A Battery EMU for instance), which means the EMU can use the normal overhead power where its available and its local supply where its not.
    So could be useful in some circumstances (e.g. Pukekohe services – avoiding the need for electrification of that line anytime soon). But again CAPEX costs mean they won’t displace DMUs so easily yet.
    And I expect the same would apply for PHEV buses and/or Hybrid/Battery light rail options too.

    1. @12,000k per year, Toyota say about 22 years (180,000 miles or 280,000 km)
      Then again, I do about 25,000 km / year, so I’d get 1/2 that.

      But instead of wireless induction, why not TEG paint (“solar paint” – actually Thermo Electric Generation) for the car?

      No need to plug in, no special infrastructure required.

      I’m hanging out for that for my house.

  3. YEs RUC is probbaly the way future road user harging will happen (esp with newer cars in the fleet becomeing more fuel efficeint)
    Although the current RUC charge light-diesel cars is exactly the same for all vehicles SUV vs city car, there is no distinction below 3000kg,

    There needs to be a cheaper rate for light vehicles, as it is certainly clear that a Smart Fourtwo CDi has less road impact than a Mitsi Pajaro, although both would pay the same per Km RUC

  4. Presumably this will all be made irrelevant by the introduction of driverless cars, which will ultimately remove the whole concept of owning a car, and therefore change the economic model. i.e. if cars can drive themselves, it becomes crazy to own one and have it sitting idle while you work or sleep. We could all just share a far smaller fleet of communal cars.

    1. We could do that already nick, but don’t. Doesn’t have to be driverless to share, which suggests that driverless won’t change anything in that regard.

      1. How driverless will improve sharing (off the top of my head)
        i. one more passenger seat in the car
        ii. sharing costs reduced by no need to fund a park
        iii. car comes to you rather you walking to it’s parking spot
        iv. don’t need a licensed driver, so good for old, young and drunk to make trips in 😉
        v. won’t kill innocent pedestrians and cyclists, reducing insurance premiums, and costs.
        vi. lots of data generated, enabling performance optimisation.

  5. So if the cost of batteries decreases enough and the tax payer gives a generous donation these cars still dont make sense. Maybe Ford are on to something bringing back the XR8 next year, a 5.0 litre supercharged V8. It will have its own micro-climate of greenhouse gas.

    1. Let me fix that for you; as the cost of batteries goes down, which they will as the supply chain ramps up, and the cost of petrol goes up, which it will, as supply and demand are clearly on a knife edge despite the Shale boomlet, then these things will become more viable.

      But still it will be those for whom capital cost is unimportant, ie those who already pay silly money for their cars, that will buy the luxury PHEVs [BMW/Lexus etc] and BEVs [Tesla] and dodge contributing to the NLTF, and the capital constrained who will be stuck in inefficient old gas-guslers, paying off the uneconomic PPP RoNS…. So it goes.

      There will only be real choice when it becomes viable to be able to choose not to have to drive, at least not all the time and for all journeys.

  6. John are you missing the elephant in the room here? Fossil fuels are simply not a long term option any more. How long term? You would think well within the timescale of your projections.. Which go to 2040.

    At present rates, by the mid 2030s we will be well past any concept of a “safe level” of atmospheric CO2.. not forgetting that there is a substantial body of well informed scientific opinion that is of the view that a safe level is (was) 350 ppm, and we are already passed 400 ppm.

    Sanity must surely prevail, and it’s hard to see how without major shifts in pricing and/or availability of fossil fuel based personal transport (driverless or otherwise)..

    Whether that translates to people driving BEVs or not is an interesting question. But the future for FF powered cars is closing.

    1. Agreed. The research I’ve done into EVs is what has led me to conclude that we (and countries around the world) need to put a heck of a lot more effort into public and active transport to reduce transport GHG emissions. More on that in future posts.

    2. Interestingly China is reducing pollution and reliance on fossil fuels by mandating that 30% of all State Vehicles be alternative fuels by 2016. Hopefully a trend that will be taken up by other countries.

    3. Yeah, but it might still make sense to generate “fossil fuel” by renewable means instead of using batteries. Depending on energy density.
      Remember that Boston Legal episode about the environmental cost of making hybrid batteries 🙂

  7. Once charging technology improves, a lot of this will become redundant. Make things in large enough quantities and the prices come down as well – large lithium ion batteries are no exception.

  8. I’d love to hear what the actual lifetime of batteries has been in NZ for hybrids like Toyota Prius and Honda Insight.

    Those have been around long enough to see whether the initial 8 year estimates (that I had heard at their introduction) was pessimistic or optimistic. While Hybrids exercise batteries differently to electric only vehicles, they must be an indicator.

    1. I think those batteries have generally performed OK, and just as importantly they’ve been fairly cheap to replace when it does come time for that. However, modern EVs are using quite different batteries. The old hybrids tended to use NiMH, and all the new cars coming out are using lithium-ion instead, so the results from the old batteries aren’t really that relevant. It’ll be some time before we get a better handle on the new batteries, but for now, manufacturers are often offering fairly long-term warranties to mitigate the risk to the buyer.

  9. Nice work John, there is not a lot of TCO data on EVs. The biggest driver of the EV market will be the price of petrol. As more hit the market and more options appear then uptake will increase. Currently a NZ Nissan Leaf will cost $39,990. Nissan are now offering lease battery options where you purchase the vehicle (75% of cost) and lease the battery.

  10. Thanks John,
    Do you have any figures for the cost of the batteries or the measured life? I understood LiOn batteries are more expensive than NiMH. I what the resale value of 5 or 6 year old electric vehicle is will be compared to the cost of a similar and similarly aged petrol vehicle (for now, leaving aside the fact that we are running out of fuel).
    I ask because I want to know. I think electric vehicles are very exciting.

    1. Hi Rob, I don’t have accurate, current battery costs, although as I wrote above, they’re likely to be below $500/kWh. That’s for lithium ion. NiMH are cheaper, but not really viable for electric vehicles – the batteries are too heavy and take up too much space, relative to the amount of energy they can store (and even lithium ion could have a long way to go in that regard). As for the other questions, they’re good ones but I don’t have a good answer sorry!

  11. John, your conclusions relate to EVs at current battery prices, right? According to your graphs and others I’ve seen, we can expect half the current price in 4 to 6 years (16 to 11% per year price reductions). Surely then it will be no contest? That’s not very far away- “when I buy my next car” for most Kiwis?

    1. Hi NCD, predicting the future is always a tricky business… but yes, it probably won’t be too long until PHEVs are cost competitive with *new* cars, under the current transport charging system and for drivers who are likely to do quite high mileage (e.g. company cars). If we go to universal RUCs, it could be a bit longer. BEVs will take a bit longer than PHEVs to become competitive.
      Most cars in NZ, though, are imported second hand, and it will be years more before we start to get any EVs in that category, and it’s hard to know what the prices for those will look like.

  12. To date I have found your blogs to be very good, John; unemotive, data-driven and with a logical flow. I was looking forward to your offerings on electric vehicles with some anticipation. My hopes, however, have been cruelly dashed:

    “Today, I’m looking at the costs of these cars – both their running costs, and their capital costs”
    You may have looked at their capital costs but you neglected to tell us what they are. The information is out there and not hard to find.

    “These cars are much more expensive than conventional cars”
    That’s an assertion. I’m not disagreeing but where’s the data?

    You have used a figure of 8% for your discount rate. Why? I appreciate that there will be a variety of rates dependent on the individual’s financial circumstances but why did you choose 8%?
    You have used a figure of 12000 km pa for the purposes of your analysis. Why? Your stated extreme case is far from extreme in respect of annual km and discount rate (if I were to consider replacing my wife’s car for example the figures would be 42000 km and maybe 2.5% respectively).
    You have used a petrol price inflation figure of 1.3% for your extreme case. This seems to me to be anything but extreme. Seems overly optimistic to me but perhaps you have some credible projections…

    “At current price levels, BEVs have running costs that are only marginally lower than petrol-electric PHEVs”
    Unless I am mistaken that’s not what your graph is showing. Neither is presently subject to RUCs and your graph is showing a ratio of more than 1 to 2 in favour of BEVs but then those estimates are only as good as your assumptions re the petrol/electricity mix of the reference PHEV.

    “Overall, if you compare these running cost savings to the extra capital cost, it looks like the financial argument for BEVs and PHEVs isn’t quite there yet.”
    So you say…I have inserted my figures into your spreadsheet and used actual capex figures for several actual BEVs and concluded the opposite.

    “Since EVs also contribute to road wear and tear”
    In the context of the 4th power law and heavy vehicles it is negligible. While RUCs (in their present form) may seem logical, viewed in the context of a driver who seldom sets wheel on a state highway they are not.

    I will conclude with a somewhat rhetorical question: since when was the selection of a vehicle made on purely economic grounds?

    1. Hi MFD, sorry for dashing your hopes! Please be assured it was inadvertent and with no cruelty aforethought. This post, and most of the rest of the series on EVs, is paraphrasing stuff from my thesis, linked somewhere above, which contains a bit more detail on some of these points. I’ve kept the post light on data and pitched it as a bit of a worked example. It’s a bit hard to compare prices between BEVs and conventional cars – the price differentials seem to range from country to country, even ignoring subsidies and so on, production is still fairly small scale, etc. It’ll become easier to make comparisons as more models of BEV and PHEV become available; it’s interesting, for example, to see the relative success of the Outlander PHEV, which is can be more easily compared with the conventional Outlander.
      The 8% discount rate is given a bit more background in the thesis. I’ve looked at it as being a reasonable ballpark for a theoretical rational consumer, or perhaps a business, or indeed the government, who is trying to work out whether the purchase of an EV makes financial sense in the long run, compared to a “new” conventional car and ignoring any effects of selling the car to someone else down the track. At 8%, you could argue that if the financials stack up, but consumers aren’t buying, then perhaps there’s a market failure and a case for govt intervention, again in the limited case of EVs vs conventional new cars. Considering second hand cars, the cost effectiveness of doing so vs other potential interventions such as PT etc, makes it more complicated.

      Oops I had more to write but am out of time for tonight, will continue tomorrow!

      1. Thanks for the response, John. My hopes are now un-dashed. I figured that you had used your thesis as the basis of the discussion. Things are moving fast with respect to BEVs in NZ and I suspect that some of the assumptions used in your thesis can now be substituted with actual data. I am currently viewing the analysis from the standpoint of a rational (I think) consumer about to buy a BEV. I don’t pretend that my circumstances are typical or average. I have 3 overall objectives to meet: reduce household costs, do something significant about our household GHG outputs, have some fun tinkering with technology that doesn’t involve sitting at a computer. A key condition in meeting these objectives is that my wife must be kept happy.

        A Pareto analysis of household expenses has identified education as the single biggest expense with transport associated with that education being the second biggest. Dispensing with private education for the children would result in a big decrease in both expenses but fails the “keep the wife happy” condition. Turning, then, to the transport associated with that education we can rule out PT as there is none for 30 km. The school offers buses but the charge of around $2k pa per child, the fact that it only gets to within 8 km of where we live and takes nearly 2 hours out of each day (since the catchment area is so large) and a car is still required. Car pooling with a neighbour was tried and failed when the neighbour’s circumstances changed. BEV a potential solution to be explored.

        GHG emissions. Analysis of the source of NZ’s electricity, marginal generation type and the fact that I have 3 kW of PV to install indicates that a BEV will make a big dent in our household emissions. 30 000 km pa in a Honda Odyssey’s worth.

        Have some fun tinkering captures the fun elements of a better driving experience of BEV vs IC, exploring a product concept relating to BEVs, hooking up the PV and using the whole thing as an educational opportunity with the children.

        Turning to the figures:
        I am looking at a discount rate of around 2.5%. With no borrowings and significant savings the reference point is a term deposit. Accounting for inflation and the fact that I am taxed at the headline interest rate at the highest tax rate.
        In terms of how many km per year can be moved from IC to BEV it’s looking like 30 000.

        Why BEV and not PHEV? Decades of experience in engineering makes me a firm believer in the KISS principle (keep it simple, stupid). PHEVs have a lot more stuff to go wrong or maintain and a look at our big vehicle maintenance expenses over the last few years points to transmissions and cooling systems. BEVs have neither. Electronics aside, the BEVs I am considering are mechanically very simple.

        I have almost settled on a Nissan Leaf. We don’t have a lot of electric vehicles to choose from in NZ. It’s about the right size, the safety features are good the range is adequate, it’s pleasant to drive and it has a good track record outside NZ. It has almost satisfied the “keep the wife happy” condition. The remaining issue to resolve is whether to buy new at $40k or used (10 000 km) at $25k. A Trademe search on Nissan Leaf will give new and used prices. I have a deep aversion to buying new on the basis of the depreciation rate in the first minutes of ownership and the warranty conditions of a new Leaf will preclude my using it as a test bed for the product concept. On the flip side the lack of a battery warranty on a used Leaf is a concern. I could just wait until the present NZ new Leafs are on the used market but that would miss some of the monetary and GHG benefits.

        That’s one persons rationale for choosing to go the BEV route. Feel free to tear it apart; it may help me with my thinking. From my standpoint a BEV seems fit for purpose and I have defined my purposes. I don’t require it to render petrol or diesel vehicles obsolete, I don’t require it to solve any congestion problems (we don’t experience that) and it is not solely a choice based on economics…but at the end of the day the most important thing is to keep my wife happy.

        1. I have been working on EV TCO
          2014 Nissan Leaf 30,000km per year over 10 years would save 47% TCO compared to a 2014 Mitsubishi Lancer GSR
          2014 Nissan Leaf 15,000km per year over 10 years would save 29% TCO compared to a 2014 Mitsubishi Lancer GSR

          1. …and they say that EVs are not economic.

            That being said your TCO figures are estimates based on assumptions of future cashflows. It would be interesting to see your assumptions.

        2. With 30,000 km a year, presumably in a fairly regular pattern i.e. staying within the car’s range limit, you sound like a perfect candidate for shifting to a BEV, MFD. The opex savings will add up, you’re not paying RUCs for the time being at least, and you’ll get the most out of a battery warranty if you’re buying new (haven’t checked for a while, but is it something like 8 years/ 200,000 km, which you’d do in about the same time anyway?). You’ll know more than me about whether new or used is the best option… you can probably find some figures on whether there are many Leafs getting batteries returned.
          Back to your earlier comment, it was probably a bit lazy of me to say “At current price levels, BEVs have running costs that are only marginally lower than petrol-electric PHEVs”, but I was thinking of the RUC exemption as being something that can’t be relied on to remain in the long term – it does have an expiry, at which point the govt may decide to renew it (policy brought in under Labour, already renewed under National) or do something else.
          12,000 km is more or less the average distance driven by NZ cars, which is why I used that figure, but of course you’ll get better value out of an electric if you do more than that (and newer cars, on average, tend to do more travel, so again I wasn’t factoring that in).
          And yes the data and assumptions in my thesis are a couple of years now so there’s certainly room to bring in some real-world ones!

Leave a Reply