I finally got in the hacking mood and made up a little opto-flective sensor for my gas meter.

A few lines of code in an Arduino and I have it pretending to be a CurrentCost CC128 squirting out data into Dale Lane’s CC GUI.

I’ve done a little write up in my blog

http://sustburbia.blogspot.com/2009/11/using-arduino-to-monitor-gas.html

Hope to see you at Homecamp 03 – when they finally choose a date. Looking forward to the “Rise of the web connected sensors” and the”Web 2.0 – the internet of things”. It’s not quite in the Terminator class – but a start. Hope to be unveiling my “Cheapernet Gateway” the poorman’s plug into the internet for real time renewable energy monitoring.

Out of interest, what kettle do you use? We bought an eco one from Kenwood and it is the least user friendly piece of kit I’ve ever had the misfortune to buy. Trying to fill it requires a crowbar, and it pours badly. Thanks.

It sounds like that you have software and web skills whereas old fogies like me (44) are happier with a hardware solution.

To follow the regulations to the letter, any device within the meter zone would to have Ex certification – and I suspect that not many £50 web-cams meet that criteria.

A snapshot every 15 minutes might be OK for trending but only a pulse counter that can measure the on and off times of the pulse will tell you whether your boiler is running at full power or modulating down. Timing between pulses is also a means of spotting which of your gas appliances is using the gas – whether boiler, cooker, oven etc. A handy way of spotting pilot light consumption and any gas leaks.

For energy monitoring to truly work, the gas companies would have to roll out new gas meters for almost all homes – a cost they don’t want to bear. It has been suggested that gas meters would be monitored with an attachment that kicks the data back to the smart electricity meter – which at least has the power supply and comms link available to get the data out.

I’ll look forward to a demonstration of your webcam/FTP/OCR at a future HomeCamp

Ken

1) Get an IP camera with InfraRed vision
2) Stick it in front of your gas meter,
3) Have it automatically FTP an image of the meter reading every 15 minutes (or whatever level of precision you want).
4) Run OCR on the images
5) Or simply display the live stream of your meter whirling away.

Hey presto – an ersatz meter reader which doesn’t interfere with the gas equipment.

I got an all-singing all-dancing camera for £150, but a basic WiFi model is usually under £50.

T

The regulations are stringent and the gas industry takes them seriously. They treat the whole volume around the meter as a potentially explosive atmosphere, where there must be no source of spark, ignition or overheating component that could potentially cause ignition.

How they can enforce this in properties where the gas and electricity meters share the same cupboard and illuminated with a 100W bub – I don’t know!

Needless to say the hardware to read a pulse output is trivial – the cost of regulatory approvals is not.

I recently found some 1mm fibre optic cable that can be purchased from the likes of Farnell. It would be possible to make an opto sensing device that used several metres of fibre optic cable to convey the LED light to the reflective target and then back to a remotely located photodiode thus keeping all electric wiring well away from the meter.

If you are fortunate enough to have an RJ11 connector on the bottom of your meter, then a LED and a couple of duracells or a 3V coincell and a suitable series resistor would make a 1 way optical output.

The thermocouple method would need to have thermocouples on both the flow and return pipes to the boiler and preferably with a flow meter fitted inline – so that the temperature difference and the flow rate of water could be measured to give a more accurate figure for boiler output power and thus gas consumed.

I found that those decorative gas fires are extremely wasteful in gas, and a radiant gas fire or a boiler with pilot light can use 6kWh a day in gas on the pilot alone!

If it was you that spoke, I remember your talk on gas and it was inspiring.

I am talking to the CTO of GEO about pulse monitoring, but I think your point about stringent safety regulations will apply. My last inspection the guy from Southern Electric got really upset because I have a power and ethernet cable running across my main gas feed pipe!!! (about 2 metres from the meter no-less).

I could always hack it together myself, and may well do so (I have a couple of the Current Cost pulse counters) but I would really like it integrated with the GEO.

I am really intrigued about the leave-on-all-the-time, vs morning blast methods. My house is 1970s insulated walls and brick internal walls, so will be interesting to see.

The gas monitoring GEO use is a thermocouple on the output of the boiler (as terence says), thereby monitoring hot water and heating in my house, so will be a reasonable measure I think, and hopefully enough for me to make a comparison. The only unmetered heating variable is the lovely ‘pretty’ gas fire my wife likes to use in the lounge…

Thanks for the update. I heard you speak at HomeCamp 09, and I am pleased to hear that the kit you have installed is helping you make energy savings.

The problem with gasmeters is that they are all different, and only the more modern ones have any facility for a pulse output. As you have an RJ11 (R5 port) on your meter, that’s half the battle.

It would be relatively simple to get an Arduino or similar to count the pulses and get them up to the net using an Ethernet shield.

However, the gas utility generally frown on any suspicious devices connected to their meters, and from a legal point of view, any equipment connected to the meter has to be shown to be intrinsically safe, and meet quite stringent safety regulaltions.

In the late 1990s, I developed a wireless pulse counter which sent out the accumulated pulse count out on a simple low power 433MHz link. It was battery powerered from a lithium AA cell, but certified that it could not develop a fault which could ignite a gas/air mixture or lead to an ignition temperature. That was the real challenge and the design had to be tested and certified by BASEEFA – a costly exercise, even for a small company let alone an individual.

IMHO its the safety regulatory requirements surrounding gas meter devices that is restricting the development of gas monitoring solutions.

As a second point, regarding your comment for gas consumption for heating a typical house, is it cheaper to keep the heating on all day 24/7 at a low setting or have it come on twice a day?

This I believe is entirely down to the construction (ie age) of the property. I have found that with my 1905 brick-built semi, with solid 9″ walls that it is cheaper and more comfortable to keep the heating on 24/7 in winter, at an approximate power of 6kW, than to blast it for a couple of hours at 24kW in the morning and 4 hours in the evening. This is solely down to the high thermal capacity of the walls which need to be brought up to temperature and then maintained so.

For a modern, well insulated house with low thermal mass plasterboard partition walls, the converse is likely to be true. As I see it, you only have to heat the air in the room and then the insulation helps maintain the room at temperature.

Using this approach, I managed to shave off 6000kWh of gas per year, bringing my 19000kWh bill down to 13000.

Whilst this might appear contentious, it’s what I have found after 9 years of energy monitoring in this property.

I was always told that you should keep the hot water tank heated at all times – it’s better to keep it hot all day than boil a whole tank every morning / evening. I guess it depends on how efficient your boiler is and how well insulated your tank is.

I’m getting a little obsessed with my GEO Duet. I like your idea of showing cost rather than kWh.

The other thing I recommend – to get your costs down – is switch energy provider. Your current one should be able to tell you how many kWh you use in gas and electricity. I’ve move from a tarrif which would have cost £1,200 to one which should cost ~£850 (assuming price and usage stays the same).