Friday 30 March 2012

Solar PV - the basics

As I mentioned in the original post, as well as solar water heating we also have (as of last week) solar electricity, or solar PV as it's more commonly known.

I thought I'd start out with a brief description of how a solar PV system goes together. First, a diagram!

A generic domestic solar PV system
Ok, So...

The two main components of the system are the PV cells themselves, and the inverter.

PV is short for "photo-voltaic' from the Greek photos, light, and Alessandro Volta, who invented the first battery - basically PV cells turn light into DC electricity, and the more light, the higher the voltage they generate.

Now, DC is not much use for running household equipment, as that requires 240V AC. Enter the inverter, whose job, pretty simply, is to turn DC into AC. This is where things get a little less simple, since inverters have a minimum voltage at which they will 'kick in', so if it's not bright enough to generate that minimum voltage? No electricity. Once it does kick in, the higher the voltage off the PV cells, the more power (and thus the more units of electricity) the inverter kicks out.

In our configuration[*], the inverter feeds into a spare breaker on our fuse box, via a meter that measures the total units of electricity generated (very important, as this is how you get your Feed-In Tariff, of which more in a later article). If there's more being generated than the house is using, then (by the magic of electricity) the excess is fed back to the grid, otherwise the grid supplies the balance.

And that's pretty much it. Other than to note that, if you were dreaming of being able to go 'off-grid' when the power goes out? Most systems don't, and more awkwardly, they actually shut down when the power goes off. Annoying and somewhat paradoxical - the reason behind it is that in order not to cause disruption to the grid, it's essential that the inverter syncs up to the frequency of the grid supply, and it's a requirement that if the grid goes off, it shuts down until it comes back.

Next in this series of posts, I'll look at how to calculate what kind of power you can get from a PV system.

[*] In some systems, the inverter is connected up in the meter cabinet instead of to the fuse box: this makes little or no difference to how things work.

Tuesday 27 March 2012

E.On's Energy Fit monitor

We've been E.On customers for a while now, and in around 2010 they sent all their customers what they called their "Energy Fit" Starter Pack. In a nutshell, it's a clamp sensor, plus a small wireless transmitter for it to plug into, a base unit with a big LCD, a USB cable, and some documentation and a CD.

I was not at all surprised to discover the CD was Windows-only, and had a gentle go at E.On for that. Reading around a bit, I discovered a couple of useful things:
A bit more Googling around revealed the CurrentCost page with the specs on how to talk to it, and a few assorted folks' sample code to do so, and the fact that the USB cable was simply a USB-to-serial adaptor. Brilliant.

Not that much coding later, I had a Perl script on our Linux server (which I'm not going to share just yet, for reasons I'll explain in a later post) that both displayed the instantaneous value of the clamp sensor on a web page, and uploaded the data to Google PowerMeter. Which was fab, right up to the point where Google decided that, and I quote:

 "...our efforts have not scaled as quickly as we would like, so we are retiring the service." 

Colour me unimpressed. Especially when earlier in the same press release they said, quote:

"Studies show that having simple access to such information helps consumers reduce their energy use by up to 15%; of course, even broader access to this information could help reduce energy use worldwide."

Lovely. Nice one, Google. Way to help out there. Nice to see the studies back up my views on monitoring, though.

The script's still running and saving the data, and I'm currently feeding it into Splunk, which is a sysadmin tool I've used at work, and comes with a free version which'll handle the volume of data we generate. While it's an über-geeky solution, I wouldn't recommend it since Splunk does require a pile of customisation to generate wife-friendly dashboards, and there's still the overhead of running the background script to feed it. 

But - it has taught me a fair bit about what it's capable of, and a lot more about what else was out there. If you can get your hands on one, be it the E.On branded one or a CurrentCost, they're pretty easy to set up and get info out of, and you will learn a lot about your power usage habits.

Monday 26 March 2012

Why monitor?

It may seem an obvious question, I know. And I'd be the first to admit that having devices I can interrogate remotely appeals to my inner geek, as does making stuff that isn't supposed to talk to anything but a closed-source, badly-written Windows app or a proprietary web service bend to my will. I'll freely confess that I'm a gadget freak, and that generating our own hot water and electricity is just cool.

But this isn't just about being geeky: it's about knowing what's happening in your system. For example, since having the solar water heating, I'd bet that we've saved nearly as much through being aware of what hot water we have and how long it takes to heat, as we have by generating it ourselves.

Equally, since I got the clamp sensor installed and talking to the computer, all of us at the Mill House have been very aware of saving energy - if you know that (for example) your house with most of its 'always on' stuff (fridge, freezer, server, etc) ticks over at 380W, then if you see a reading of 700W you'll start to investigate what's causing it. At some point I will let Anne post a breakdown of last year's energy savings here, but for now I'd just note that our electricity usage dropped by around 30% over 2011, just through knowing what we were using. (And, I should add before Anne does, through finally shutting down two Linux servers with very old power supplies. Old power supplies become very energy inefficient (several hundred Watts worth), which is why our house server is now a 60W miniITX box from the good folks at along with a similarly rated HP Media Server as a fileserver.)

But of course, that's just the first step. Sure, you can install your solar PV and sit back and watch the Feed-In Tariff money roll in on sunny days. But shouldn't you be using all that green electricity? As an example, our dryer is now on a timer (and once I get it set up, it'll be on a computer controlled switch, in fact) that only lets it run at times when the PV is likely to be generating most if not all of the power. (And sure, yes, we could hang it out in the garden. But both of us work, so being in to remove it when it rains isn't so easy...!)

Given the tightness of everyone's budgets generally, let alone the current move towards being at the very least energy efficient, and ideally as green as possible, I personally think it's vital to know both where your energy is coming from, and what's using it, long before the bill turns up. And setting things up to do so doesn't cost that much.

Clamp sensors

Currently, our electricity supplier is E.On (though we're looking at Ecotricity) and, back in 2010, they sent us a free energy meter together with a clamp sensor. More on the meter later, but for now I'll briefly discuss the clamp sensor, and what it does and doesn't do.

In a nutshell, what you do is clip it round the incoming feed to your electricity meter (or, in fact, any current-carrying cable), and it has a go at measuring the power consumption through the cable. What the meter connected to it then does is work out the current in your cable based on what's induced in the clamp loop, and then make a bunch of assumptions to calculate the power. Specifically, unless you have a smarter than average device at the other end, it assumes it's seeing 240V (in the UK), and then

power P = voltage V * current I * a fudge factor.

This falls short of ideal in a number of ways:
  • your mains voltage can fluctuate a bit
  • the actual current induced in the clamp can be affected by precisely how you place the clamp
  • it takes no account of the kind of devices connected to your electricity
This latter can be quite noticeable. I won't go into details, but if a device has inductance or capacitance as well as resistance, then the equation isn't just P = V * I, and it depends on the nature of the particular device. Things with heating coils and big motors tend to be particular culprits, and most of us have one or more of those in the house!

There's also another problem which rears its head when you install solar PV cells - when your inverter kicks out more power than your house uses, the rest gets fed back to the grid. Because of the way it works, the clamp sensor can't measure which way the current flows, so it can't tell if you're using or generating power. 

So why use a clamp sensor? Well:
  • it works on any current-carrying cable
  • if you have an old-style electricity meter, it's your only choice
  • if you have solar PV, unless you have a smart meter that your PC can talk to, it's your only way of measuring how much electricity you send back to the grid (in conjunction with another sensor)
The latter is why we're hanging on to ours. More on our monitoring system in later posts.

Solar water cost and ROI

Our panels (a pair of evacuated tube panels) set us back of the order of £9000 in (if I remember correctly) 2004. With hindsight, I think we did this too early - the price of a similar system has now dropped considerably (more like £2500-5000, apparently). We're probably saving £50-80 a year, so the ROI time is... a while yet.

On which topic, it's worth noting that our suppliers went out of business, and were taken to court indirectly over their rather inflated ROI claims, and we did get £4000 back as a result. We're currently using the local Peterborough Boiler Services to maintain our system (not that it needs much!), and are pretty happy with them.

It's also worth noting that there are two government incentive payments for solar hot water, one of which is available now, one to come. The one available now is the Renewable Heat Premium Payment, which is (or pretty much was since the deadline is March 31st) a scheme which pays £300 for any solar water heating install since Aug 1st 2011. The future one is the Renewable Heat Incentive - basically, if you've installed since 15th July 2009 (bother!), you will be eligible for some form of payment similar to the solar PV Feed-In Tariff.

As I said, with hindsight, we should probably have left it a couple of years: I'll get some stats upon a subsequent post for the last few years' gas bills, so you can see how much we've saved. Either that or I'll give Anne posting rights on this blog, and she can!

Our solar water heating system

Our solar water heating consists of a couple of panels on the roof of our outbuildings. They show up very clearly in the Google Maps aerial view, although (as of this post) the Maps view is out of date, as we had them moved in 2010 one 'bay' of the roof closer to the house.

The control unit is mounted on our inner landing next to the airing cupboard. Very usefully, it comes with three temperature sensors, one monitoring the temperature of the panels, the other a pair that measure the temperature at the top and bottom of the hot water tank. Essentially, how it works is that if the temperature of the panels is greater than the bottom of the tank (and less than 60ºC), water is pumped round the panels, heating it up. [The astute will note that this uses electricity. The really astute will note that this uses electricity when the sun is out, and we (as mentioned) have solar PV for that!]

Since we've had it installed, we pretty much don't pay for hot water between mid-March and late October, as the solar heating is generally enough to maintain enough of the tank at 40ºC or better for baths, showers and washing up. Sadly, it's not possible to buy a hot-fill washing machine any more, though, so ours does heat water itself for laundry.

I'd go so far as to suggest that even if you don't get solar water heating installed, sticking a pair of contact temperature sensors on the top and bottom of your hot water tank that you can monitor is incredibly educational. Apart from anything else, it cures you completely of setting the central heating timer to heat your water twice a day for an hour, since:
  1. we certainly don't use that much hot water in a day
  2. the boiler will heat the top of the tank to 60ºC in about 20 minutes flat (and remember, the top of the tank is where you take water from!)
Sadly, our controller doesn't talk to a PC, so as of yet we don't graph the temperatures or track if the pump is running. As ever, however, watch this space, as I have plans.

Welcome to the Mill House

The Mill House is a 19th century stone-built house in Werrington, just north of the centre of Peterborough. It was built by the miller of Werrington (one Mearston Oldham) in the 1860s and 70s, being right next to the mill (in fact, our outbuildings connect).

Since we moved in in 2002, we've added solar water heating, solar PV panels and a loft extension, and the aim of this blog is to record some of our experiences with installation, use and monitoring of the systems.

Some bits of this blog will be a bit geeky - I'm a Perl programmer by trade, and quite prepared to use my job skills to produce useful data on our systems. One of my medium-to-long-term goals, though, is to produce a boot SD card for a Raspberry Pi that allows anyone to plug a few monitoring gadgets in via USB and Bluetooth and display the results on their TV.

So... watch this space.