I have had a fascinating, but ultimately frustrating, discussion over the past 24hrs about the reliability of the Ecotricity Electric Highway. Here I expand a little on a few of the basic premises and assertions I make regarding current reliability.
The first point of discussion surrounded the issue of whether the network reliability changes as more people use it?
I asserted that no, the reliability of the network does not change in any way as more or less people use it. The reliability relates to the risk a charger will be failed on any particular trip.
Let me explain…
Assume there are 10 locations in the charge network and 2 of them are not working. Let us also assume that these chargers are used 100 times a day. Then if I set out on a random journey then I have a 2:10 chance of coming across a non-working charger and being stranded.
Let us assume now that instead of 100 uses a day that there are 500 uses a day, how does that affect the chances that I will come across a non-working charger? Well it doesn’t. My chances of coming across a non-working charger is the same at 2:10. OK, I might have a bit of a wait as there are more people using it but the reliability is still 2:10
This is the basis of my determination of reliability as perceived by the driver. The risk of being stranded is related only to the ratio of working chargers to non-working chargers and is not at all related in any way to the number of people that use the chargers.
People seem not the grasp this basic premise.
So I say here and now… the risk of being stranded, and hence the perceived degree of reliability from the drivers perspective on a single journey, is not affected by the number of people that use the network. The only data that is relevant is working and non-working chargers.
If you think I have missed something then please say so in the comments.
It is on the basis of this basic premise that I post the figures regarding non-working chargers. I want the figures to show a 100% working and 0% non-working but they don’t and the ratio is not showing significant signs of improving and while that is the case I want everyone, particularly people assessing the state of the Highway in view of buying an EV, to know the facts. there are far too many EV evangelists prepared only to publicise the good bits. I want people to know about the good bits but also the not so good so they can make an informed decision.
Currently, as of this morning, the ratio of non-working chargers to total chargers is 35:247 or a 14.17% of the network is not working. It follows then that I have a 35 in 247, or about a 1/7 chance of coming across a non-working charger on a random trip using a single charger. The risk goes up as more chargers are used on the trip.
I don’t call that a reliable network!!!
The second issue was that it was suggested that as we are not seeing many people stranded so that was an indication that the network was inherently reliable. I disagreed.
Currently, the vast majority of people know that the network is unreliable and so before they travel they check on-line to see if the charger they intend to use is working. If it is great. If not they look for alternatives. Sometimes they will find working alternatives and use those. At other times there may be no alternative so they have no option than to cancel the trip or use an ICE.
My assertion is that this advanced planning significantly skews the figures and so we cannot use the charging and stranding figures in any way as an indicator of reliability. If people are planning ahead and so change their plans so they are not stranded, or cancel their trip again so they don’t charge at all and so are not stranded so the figures surrounding the number of charges made or the number of strandings do not represent the actual risk.
It is good that people are not stranded. But the fact that so much pre-planning was necessary to arrive at those charging and stranding figures means that the data is highly biased to today’s way of planning trips. I assert that public acceptance of EVs will only come about if this degree of planning is not necessary (like with petrol cars today) in which case today’s figures are not relevant. We must assume that people are not planning in advance to this extent and so I say that we must consider today’s stats as irrelevant.
It has been suggested that if a charger is never used then the fact that it is not working has little impact on the network. The example given was Land’s End and Oxford. Clearly Oxford will have more traffic than Land’s End and so they suggested that the importance of these two locations, and hence their impact on reliability, was not equal.
I can see this argument if you want stats that cover reliability from a network point of view such as the network operator might want. However, to a single driver, on a single trip, how many others have charged at that chargers today has no bearing at all on the impact that a failed charger has when you try to charge at either and fail. OK, it won’t affect as many people at Land’s End when compared to Oxford but the chances of coming across a failed charger is the same. In fact, it probably affects the Land’s End driver more when it happens as there are less alternatives. However, keeping it on a simple level, it is fair to say that the chances of of finding each charger working or not working are equal.
So we cannot use any stats that includes charging numbers or numbers of strandings when assessing network reliability from the driver’s perspective and the fact that there are so few strandings might indicate good pre-planning by drivers rather than good network reliability.
I will continue to fight this particular fight while the network poor reliability make travel risky.