Letters to the editor: volume 16, issue 8

Vehicle use data needs to be realistic

The conclusions reported in the story ‘Most UK cars only driven for one hour a day leaving ample time for EV charging’ on the E&T website are misleading, given that so many properties built in the 1950s, ’60s and even ’70s did not provide off-road parking. Many that did had garages separate from the main dwelling with no possibility of providing mains electricity, making it difficult to provide EV-charging facilities without considerable expense.

This article would provide a more balanced view of the opportunities for EV parking if the percentage of time vehicles were parked could be modified to include off-road parking that has access to mains electricity. This would show how many vehicles could not have the option of being charged while parked, and how big this problem is, due to planning decisions of the past that we have to live with today.

Let’s take off the rose-coloured glasses and actually look at how this problem could be addressed rather than use simplistic measurements on vehicle usage in isolation. The provision of sufficient EV-charging facilities is a major issue with a massive cost.

I am very much in favour of EVs, but we need practical and funded solutions.

Chris Chambers MIET

By email

Replaceable batteries could boost EV adoption

Dave Neale (Letters, July 2021) ponders how the refuelling process will work when most cars are rechargeable ones and identifies two major difficulties – access to a charging outlet for cars parked on streets and the time taken to recharge.

One option to minimise these problems is the development of replaceable battery packs and recent media announcements indicate that this is under way. In the short term, an easy-to-introduce solution is required. The use of a booking system, at least for some charging points at a filling station, would seem to be a way to proceed. The ubiquitous smartphone would no doubt play a significant part in such a solution.

Michael Harding MIET

Lincoln

Downside of hydrogen transport

Unlike Dave Neale (Letters, June 2021), I don’t think it is worth considering the option of hydrogen vehicles, at least not for private cars, as a general mode of transport in future. This is partly because, when looking at the whole picture of road transport, I don’t believe that refuelling/recharging EVs on motorways will be a great challenge. Also, I believe it is likely that the adoption of a hydrogen transport option would produce a worse outcome for zero-carbon transport overall because it will require vastly more primary energy supply than the electric option, a fact I rarely, if ever, see considered.

If we went down the route of replacing petrol stations with EV charging stations and carried on using EVs in the same manner as we do fossil-fuel vehicles I would agree, this would be a huge challenge and probably unworkable, but that’s not the way EV charging works.

We won’t be putting the charging points in expanded filling stations; we just need to put them in existing car parks, which is actually the current practice, so the green fields have already been concreted over. Furthermore, if we went 100 per cent electric, we probably wouldn’t need filling stations at all, so these could be ‘rewilded’.

Most EV drivers are likely to start their long journeys with a ‘full tank’, having charged them overnight, so unless they are going on exceptionally long journeys, they probably won’t need to recharge. It is quite likely that they will need to stop for a comfort break, at which point they can plug into a high-speed charger and probably put at least 100 extra miles in the tank. When they leave, they won’t have to stop at a separate filling station to fill up, so it actually saves them a little bit of time.

The main challenge with EV charging stations is the high electrical power levels that they will require. A few dozen superchargers operating simultaneously will require several megawatts of power, far in excess of what a current motorway service station would use. However, the trick here will be to buffer the incoming supply with on-site storage to level out daily demand. The situation with a hydrogen refuelling station, producing hydrogen on-site, would be several times worse. The electricity demand per equivalent tankful would be about two-and-a-half times more and there would be substantially higher usage as people won’t be ‘filling up’ at home before travelling.

Producing hydrogen off-site has its own set of transport problems due to its challenging thermophysical characteristics. A direct swap from a petrol and diesel filling station to a hydrogen one, even allowing for the improved efficiency of fuel cell vehicles, would need roughly twice as many litres of liquid hydrogen to be delivered as used to be needed for petrol and diesel, carried in either cryogenic delivery vehicles that won’t be able to carry as many litres as a fuel tanker, or as a high-pressure gas, which would be even worse. EV charging requires a one-time upgrade to the service station’s electrical supply.

The energy demand for EVs is actually not as great as many people think it to be. If everybody in the UK drove electric cars, doing the same annual mileage as we currently do, the average increase in electricity demand would be between 25 and 30 per cent. If we all drove hydrogen fuel cell vehicles running on green hydrogen we would need about three times as much clean electricity, almost double current electricity generation. That’s a lot of extra wind turbines to be built and installed and, furthermore, the effective cost per mile would probably be about four times higher. The well-to-wheel efficiency of HFCVs is roughly one-third that of EVs, plus there’s the cost of all the electrolysers to be paid for.

Tesla has already demonstrated battery-swapping technology, which takes about half the time it does to fill a car with petrol. However, it has since dropped its development in favour of superchargers. As Dave Neale points out, it might be incredibly difficult to get an industry standard on battery swapping, although this may be an option for commercial vehicles.

With continued improvements in battery technology and cost reductions, I think it is quite likely that by 2030 we shall see affordable family EVs with a range in the region of 500 miles. If this happens there will be very few people needing to recharge on the go; even if they do, they will probably be happy with an hour or so break whilst a supercharger puts a few hundred miles back in the battery.

The issues of charging EVs are not as great as many people believe, which I think is probably because many tend to look for the difficulties rather than the solutions. Whichever option we go for it’s going to require substantial upfront investment, but once that is done the hydrogen option is always going to be far more expensive to run.

Nick Cook TMIET

Cheltenham

Why electric wins the green road race

It is disappointing to see letters in E&T that use questionable data and sources to undermine the inevitable move to electric vehicles. For example, you cannot claim that producing an EV is “no greener than a diesel until it has done 50,000 miles” only using the cost of manufacture of the car itself; the whole-life costs must be considered. A diesel car requires diesel fuel which comes at huge energy and environmental costs in terms of surveying; extraction; storage; transport by sea; more storage; refinement; yet more storage, and then local distribution by lorry. The electricity used by an EV has none of these financial and environmental overheads, only the cost of additional infrastructure (if any) to support its generation and distribution, most of which is there anyway. If you add these things into the equation then the EV wins the green race, hands down.

Another letter claims that “… road filling stations will require large areas to accommodate cars taking hours to recharge”. The latest EVs use an 800V architecture and take 18 minutes or so to charge from 10 per cent to 80 per cent. Within a few years most experts believe that full charges in five minutes will be possible. Older EVs can be recharged in the time it takes to stop and have a light lunch.

Hopefully, soon, we will see positive articles about the real benefits of EVs apart from the environmental ones. For instance, the idea of car-to-grid transfer can potentially revolutionise how the grid deals with peak demand and, at the same time, utilise currently unused overnight generation.

The introduction of electric vehicles will not be without its issues but, please, can we stick to the facts and not (mis)information supplied by the oil industry and its lobbyists?

Derek Pollard CEng FIET

By email

What happens when electric cars fail?

Two years ago, I purchased a five-year-old BMW i3 REx. It had covered 36,000 miles and was the perfect car for town commuting. As a product and package for that purpose, the i3 is a tremendous vehicle: it’s spacious, intelligent, well equipped, quick etc. It also boasts impressive green credentials. It won 'Green Car of the Year' in 2015, with zero g/km carbon dioxide emissions. At launch, BMW claimed that it would be the most sustainable car on sale.

I have had no issues with my i3 in the subsequent two years. It has seen all necessary servicing and had one identified recall in 2020. In its now seven years it has covered 43,211 miles. It is not a sports car; it makes no suggestions that it is and it has not been driven that way. It is a lovely car and I looked forward to keeping it for many more years ahead.

Last week, the car suddenly developed a ‘drivetrain’ fault. At the time, I was near home but on a minor, quiet road. The inconvenience of a breakdown is somewhat part of motoring, albeit worrying and problematic when the car is immobile. The gearbox reverted to 'Park', so rolling to a safe position was not possible. The police assisted and recovery was eventually made by the AA. The car headed straight to a BMW dealer.

Fault diagnosis strongly pointed to the electrical machine electronics (EME) unit and it needing to be replaced. The cost of said repair, although not guaranteed to be the solution, is approximately £7,800. At this level, the decision to proceed with repair or to write off the i3 is very difficult to make. Why has an electronic module (a power inverter) with no moving or wearing parts catastrophically failed after only 43,000 miles? The main battery carries a warranty for eight years/100,000 miles, so why not other similar components?

The diagnosis itself took several hours. How can a vehicle that knows enough of a fault with itself not be able to communicate the exact fault conditions under BMW interrogation? I can control the climate with my smartphone and send destinations to the satnav, but BMW technicians needed four hours of trial and error to form an opinion on what could be wrong.

How can a car be simply written off by an electrical module? How can the unit cost per module be justified given the large volume in circulation and still currently in manufacture? Why does the system architecture involve the locking of the VIN and EME to each other, thereby monopolising all aftercare to dealership only? How is it environmentally friendly to manufacture a vehicle - specifically aimed at the conscientious, caring environmentalist - that can so easily be put out of use by cost of parts and labour?

Generally, age-related degradation mechanisms are predictable and it is accepted and expected that the probability of failure increases with age. Therefore, it shall be more common for older, second-hand cars to develop and present faults. More than 200,000 i3s have been produced since the model was launched in 2014. More and more will be filtering into the second-hand car market. And herein lies the problem – an impending wave of untreatable component failures that render the resale value worthless.

I am currently waiting for a scrap valuation for my i3. In the meantime, I believe it is necessary to make people aware of the potential economic pitfalls of buying such a vehicle.

Robert Weaver

By email

Judging when it’s time to trade in

So now cars that are reliable and last a long time are bad news (The Graphic, July 2021). Give me strength! What am I to do?

I first bought a diesel quite a few years ago when I was working and driving 25,000 miles a year. The government encouraged me to buy one because the low carbon dioxide emissions were seen as environmentally friendly. My latest diesel (the fourth) was bought in 2010, about a year before I retired, when diesels were still seen as good. It was an ‘eco’ model with lots of features to make it kind to the environment, including high-pressure tyres, exhaust-gas recirculation and a carbon particle filter in the exhaust. As a further incentive, road tax was only £30 per year. It does 55mpg on short journeys and 80mpg on long ones. The modestly sized fuel tank can give me a 900-mile range. I never really thought I’d be keeping the car so long, but then diesel and petrol cars suddenly became evil and we are being encouraged to buy electric cars. What am I to buy?

All my previous cars have been second-hand or, as in the case of my present one, an ex-demo model, so were significantly cheaper than buying new. After I retired my (pre-Covid) annual mileage fell to 7,000 miles, mainly short journeys, but significantly including about three trips a year up north from where I live in Sussex, a distance of just over 300 miles - a journey that is easily accomplished with my present car with absolutely no worries about range or being able to refuel.

My present car has 107,000 miles on the clock and is running as well as ever. It is apparently worth £1,200, up £200 from last year. The cheapest electric cars are around £30,000 and have a range, they claim, of 239 miles on a long run and take an hour to recharge with a fast charger. We’d either have to sit in the car for an hour while it recharged or go into a service station and eat an expensive meal we don’t want to pass the time. I’m dubious about buying a used electric car as I hear the batteries deteriorate with use and the range is reduced.

If I took the plunge and splashed out a large sum of money, would it make any difference? I read that because of the environmental impact of manufacturing them you have to drive 50,000 miles in one before it starts being better overall than a petrol or diesel model. With my annual mileage it would be seven years before it started benefiting the environment. At my age, 79, I may have to stop driving before then. And what happens to my old car? Scrapping a perfectly useful vehicle with many years’ life in it seems a waste of resources. If it is sold on and used by someone else it will still be polluting the environment.

Denis Sharp CEng MIET

Littlehampton

Heating options for older homes

In his letter in the August 2021 issue of E&T, ‘DIY solution to green heating’, Bob Newton writes of his roof-mounted water-heating system: “When a temperature differential occurs between the collector and the ‘bottom’ of the tank, a pump in the loft circulates this ‘loop’ to utilise this ‘free’ water heating to the lower coil. In summer, the temperature of the loop far exceeds the melting point of standard solder joints in the 15mm copper pipework, so care must be taken to account for installation methods. A small expansion tank built onto the pump provides protection.”

Really? The last time I checked, the boiling point of water was 100°C and the melting point of typical lead-free solder is around 375°C. Without knowing the dimensions of his solar collector I cannot comment on his claim to have “virtually free hot water for deep baths all year with only the winter months needing to be partially supported by gas”.  

Living in a 420-year-old listed home, constructed with an ancient oak frame and single-skin brick walls, I would be very jealous of such a system. I am wondering what I will have to do for heat and hot water if and when gas boilers are outlawed. Hydrogen might work, but I doubt significant availability of ‘green hydrogen’ within my lifetime. ‘Blue hydrogen’ is only greener than natural gas if the production process is linked both to carbon capture and a scrubber process to remove all other combustion pollutants from the environment. In my mind, blue hydrogen is little better than gas produced by the ‘water gas’ process that I learned about for O-level chemistry in the 1960s.

One option for me might be to build a photovoltaic ‘power station’ in the garden, but I need to negotiate with my better half before even beginning to tackle the local planning department and its building consultants. What options do other E&T readers have?

Tim Masson

By email

Asking the right questions about machine learning

The Book Interview in the August 2021 issue of E&T asks the excellent question: “Should we trust machine learning?” Well, the obvious answer is yes. Why? Because machine learning and neural networks learn from many thousands of examples; far more than a human can entertain.

Unfortunately, though, the answer yes is to the question asked and not the one that was intended. The real question is: “Should we trust software engineers’ code to interpret and program a neural network and produce machine learning algorithms whilst ensuring all relevant data is accessed and applied in training the neural network to answer the question that was meant to be asked?” Here the answer is emphatically no. The ‘alignment problem’ has parallax: it all depends on how you look at the question.

Over the past few months, I have been involved in developing an Artificial Neural network in Machine Learning tool, ANiMAL, which is open-source and available online. This is built in Excel to predict when hand-pumped water wells in Tanzania are about to fail. Why does such a straightforward project need machine learning? Because analysis of well output alone was insufficient to ask the maintenance team to undertake the 30-hour return bus trip to fix a working well. Output was reduced for certain weeks not because a well was faulty but because the locals were harvesting rainfall so did not need water from it. Also, output appeared reduced not because a well was failing but because the monitoring system’s battery was low. This was just one other input for the neural network.

The complexity of training the neural network soon became apparent. Using the 10-neuron ANiMAL, analysis of a few inputs yielded some fantastic and counterintuitive results. However, we must be careful to ask the right question. We should not have asked, “When will the well’s output fail?” but instead, “When will the well produce insufficient water for the local community?”

No matter how fantastic machine learning appears to be, if - due to the alignment problem and parallax - we can’t even ask the right question, we certainly can’t expect an answer to the question that was meant.

Andrew WS Ainger FIET

Harpenden

Write to E&T at engtechletters@theiet.org. We reserve the right to edit letters and to use submissions in any other format.