(Ln(x))3

The everyday blog of Richard Bartle.

RSS feeds: v0.91; v1.0 (RDF); v2.0; Atom.

Previous entry. Next entry.


12:23pm on Friday, 5th March, 2010:

Underestimation

Anecdote

My mother lives alone in a bungalow with only four rooms: a bedroom, a living room, a kitchen and a bathroom. Because of this, her electricity bill only comes to something like £50 a month.

However, she hasn't had her electric meter read for ages and ages — several years, in fact. People who were supposed to come and read it never did (she got compensation from the electric company for this, twice) so her usage was estimated. She can't read her own meter, because it's in a box to which she doesn't have a key — plus she's completely untechnical so couldn't do it anyway. I opened it with a makeshift key I manufactured myself out of a bent strip of metal and did a couple of customer readings for her, but she has two tariffs (daytime and nighttime) and the meter flashes them up one after the other on a single display; it's not apparent which number is for which tariff (you have to wait a while to see which one changes, which took more time than I had available to do the reading), so I had to give both numbers and tell the electricity company I didn't know which was which. I don't know that they knew, either...

Recently, a meter-reader did appear and did read my mother's meter. The result is that her electricity bill has gone up to over £200. She's in a mad panic about this.

OK, well looking at the figures, it seems that the electricity company was underestimating her electricity consumption. They weren't doing it by much, but over the course of a several years it's accumulated. She's actually used over £500 of electricity more than she has paid for. The electricity company is spreading the repayment over 6 months, which means that her monthly bill is going to go up by close to £100 over those 6 months before it drops back down. They're flexible about this, though: she could pay the £500 off in one go, or spread it over 2 years (so her monthly bill was only £25 higher) if she preferred. I explain all this to her, and she understands, and then the moment I stop she thinks "£200 a month!" and panics again.

OK, so half that £200 or so is to catch up for electricity she's used in the past few years in excess of what she's paid for. It's annoying, especially because it's due to the poor meter-reading record of the electricity company, but once it's paid off it will go. My mother's panic will last only 6 months (yes, she can panic for that long).

What about the other £100, though?

Well, the way the electric company works, this is estimated. They look at how much electricity you used in the previous two quarters and charge you as if you were using the same amount the next quarter. Obviously there will be some discrepencies — you'll be paying more in Summer than you should be because you used more in Spring and Winter when it was colder — but on the whole these balance out (assuming, of course, that you do actually read the meter and not just keep charging people as if it were Spring/Summer, which is what seems to have happened to my mother).

So, the other half of the £200-per-month bill my mother has comes from the electricity company's estimation of her actual usage. They seem to think she'll be using £100 of electricity a month. This is twice as much as my own electricity bill; OK, so our house is gas heated, but it has computers on a lot of the time and more electrical appliances. £100 a month for my mother's house seems excessive. So how did they derive that figure?

Well, it looks as if they did it like this. They examined the two-quarters-ago (estimated) reading from August, examined the latest (accurate) reading from February, figured out how much the difference was and divided by 6 to calculate how much electricity was being used per month. They work in kilowatt hours, and because the price of electricity keeps changing you can't do a direct mapping of electricity usage to pounds, but if you do then you still get the gist of it. Anyway, this approach sounds perfectly rational until you realise that in my mother's case it includes that £500 worth of electricity which hasn't been showing up in the meter readings until now. From the electricity company's point of view, it appears that she used the £300 she paid for plus the £500 she didn't all in 6 months. Divide by 6 and that's roughly what they think she averages per month. Except, actually that £500 was built up over several years of working from unsubstantiated estimates; they should be dividing by 36 or 48 to get the true excess per month — more like £15 than £50. Her underlying usage should be £65, not £100.

As a consequence of this, the next time her meter is properly read the electricity company will discover that she has been overpaying, and will have to reduce her bill to give her the money back. If they then work on an estimate, though, they'll think she's been using less electricity than she has and the next time she'll have to overpay again. I've run the maths and this does eventually converge rather than diverge, but still, it's not exactly satisfactory.

According to my wife, who pays these things in our house, we're currently £400 in credit on our electric bill. Gawd knows how that happened, either...

Referenced by Secure Passwords.


Latest entries.

Archived entries.

About this blog.

Copyright © 2010 Richard Bartle (richard@mud.co.uk).