Erratic/Over reading Water Meter

[#1]

Ok, so this is nothing to do with Internet or Computers but I thought the Tech Help was probably the right place to post to attract attention. Knowing that there are some very knowedgeable and experienced people who frequent this forum I thought I'd relate the situation and see if anyone has any ideas.

So we bought this house in January, it was empty for 3 months while renovations were completed and we moved in at the beginning of May. The meter reading I took in Jan was 0675 m3. So the other day , I thought probably time to do another, it was 1645 m3 !! so that's nearly 1000 m3 !!. My initial thought was , that's not possible, then Oh **** I must have a leak. Then I thought can't possibly be a leak, with that amount of water, the house would be awash and the garden would be a muddy swimming pool surely.

The meter is a semi smart one, I beleive an SM100 with remote reading. It's installed internally under the kitchen sink with a stop tap before the meter. So I did some tests to see how the meter was reading, some of the time it reads normally and then suddenly it will show a rapid burst of increase. Further testing with the stop tap off showed that the erratic readings continued and showed quite large increases. Opening the kitchen tap ( feed is very close to the meter ) drained the system and then showed no further flow, so I'm pretty sure that there is no water going through the meter in this situation but the reading was still increasing! Further detailed investigations taking regular readings over the last couple of days, show that the reading appears to be stable and correct overnight but the it goes mad during the day!! According to the readings it used about 2 m3 yesterday!!

I obviously called the water supplier (Severn Trent) and told them I thought the meter was faulty and they are arranging to replace as its 10 years old and due for replacement anyway, but I can see an argument coming about the bill which if the readings are to be believed will be about £3000!.

Anyway the reason for the post was to see if anyone has any experience of anything similar or has any ideas of what might cause the erratic meter readings. I'm sure there's not a leak as :-

1 - the amount of water claimed to be used would be blindingly obvious

2 - it shows usage when the stop tap is off

3 - if it was a leak , then surely it would be continuous reading , not erratic

So anyone got any theories...

Things I've considered :-

1 - Interference from router / dect phone which are withing 2 m of meter, but they are on continously so would interfere overnight as well

2 - TV , streaming box etc , turned all those off , no effect

My latest thought is that the cable from the meter to the remote reading pad is routed outside the house where it would get the sunshine during the day. Looking at it, most of the outer insulation has deteriorated and the internal cores are visible. Heating of that cable would possibly explain why it only is erratic during the day ?

[#2]

Not had any dealings with water meters.... you say you have "observed" erratic changes on the meter when running a tap, or not... as the case may be..

I suggest you do some more testings... make a log sheet, of what you are doing.. and the test  time(of day) & (length of test) Take copious still photos.. and some "movies" on your phone, to corroborate your tests.

Can you measure, accurately, the amount of water that your run off, into , (say) a dustbin or buckets... for a specific time... (say  5  minutes) and take readings of the meter at the beginning of the run and end of run. using empty 1ltr or 2 ltr bottles. or ... if you have them. 4.5 ltr plastic container.

do the maths to work out how much "actual water" you have run off... against the "recorded" amount on the meter.

 Do the same tests, after the new meter has been installed, to collate against a new set of tests with the new meter..

You will be able to use these figures, . to back up any claim the water company has against you for your "excessive" consumption/bill

[#3]

@shutter thanks for the reply

I suggest you do some more testings... make a log sheet, of what you are doing.. and the test  time(of day) & (length of test) Take copious still photos.. and some "movies" on your phone, to corroborate your tests.

I've been recording readings every so often in a spreadsheet over the last two days. Although not measured accurately I'm pretty confident that for example, the overnight consumption of 12 to 13 ltrs the last two nights is about right. Looking at consumption before and after running a shower and washing machine is also about right.  I do also have a video of the meter when it was erratic.

Here's yesterdays table for example

Time	Meter reading	Ltrs consumed
06:00:00	1646.60879	11.3899999998921
07:45:00	1646.62446	15.6700000000001
08:10:00	1646.65043	25.9699999999157
08:25:00	1646.6665	16.070000000127
08:50:00	1646.702	35.4999999999563
09:55:00	1646.71739	15.3900000000249
12:50:00	1647.288	570.609999999988
14:15:00	1647.47413	186.130000000048
14:35:00	1647.47811	3.97999999995591
15:45:00	1648.02111	542.999999999893
16:00:00	1648.06088	39.7700000000896
16:05:00	1648.20129	140.409999999974
17:15:00	1648.40395	202.659999999923
18:50:00	1648.4382	34.2500000001564
20:10:00	1648.44957	11.3699999999426

As you can see, its fine up until late morning/early afternoon and then goes crazy for a while even though no taps on and nothing using water. It shows half a m3 sometime between 9:55 and 12:50 and again between 14:35 and 15:45. I'm sure that volume of water if leaking, I would notice! Also at that rate of flow, the sound would be apparent, no sound of water can be heard at the meter during those periods.

Having been involved in process automation during my working life, I have a grasp of how flowmeters work but not water meters in particular and I'm intrigued as to what might be causing the problem. The only thing I can see so far that correlates with those times, is that late morning is about when the sun would start to hit the cable and 19:00 is about when it would be back in shade. On the other hand, it seems strange if a simple thing like a heated cable to the remote reading pad could influence the meter reading...

[#4]

OK.. so you have the meter readings of the "assumed" consumption....  those figures are not necessarily correct. and you do need to accurately measure consumption, as I suggested, as a method of "proof" to the water company, that the meter is giving false readings.. ( assuming they differ from the actual consumption measured ).

Your "guesstimated use " for a shower, are just that.... a guess... so really bear no weight or value when in an argument with the supplier over your bill...

 One can assume that there is no consumption,  in normal households, between the hours of "say" 0100 and 0600 so you would need to do some meter readings at those times..... to confirm or deny that statement.  (unless, of course, you are running some kind of water-cooled equipment 24/7 ! )

[#5]

I'm not sure that proving its absolutely accurate when its reading normally is of much use. It's proving that its not reading correctly sometimes. I think that 500+ litres in just over a hour proves that!.

[#6]

OK.... Good luck when the bill comes in... !  ..

[#7]

One cubic metre of water weighs one tonne. 🚰

[#8]

@MisterW 

First I would try to get this witnessed, preferably by Severn Trent before they remove the meter. You don't know if it will sort itself out if it is removed to have the calibration checked. I would also be pre-empting any dispute by rattling some cages up front, starting with a formal complaint.

I assume that these electronic water meters work on a simple paddle wheel transmitting pulses to a counter principle. The ones I have seen used a magnet built into the wheel teeth to count the rotation. The problem with pulse counters is that they are bi directional. If this is the case then any stray pulses getting in to the electronics will get counted. However in real world terms these pulses will be at a slow rate, way out of the pass band of normal wireless devices. So what emits pulses at the rate of x per second? Unless you run a morse code transmitter I'm at a loss!

The other thought I had was whether the paddle wheel was oscillating forwards and backwards around the trigger point. I could believe that water hammer could cause this but you aren't going to have water hammer if the stop tap is closed.

Expanding on that thought, I suppose that if the paddle wheel happened to stop at the trigger point and there was no dead band between the pulse generator on and off trigger points, possibly due to the transmitting magnet losing its strength over time, the pulse trigger might oscillate and output a string of pulses. This could be disproved if, when you were seeing the meter clocking up at a silly rate, you turned a tap on, which  would move the paddle wheel and the rate would return to something more sane.

[#9]

And almost 220 gallons   219.97 to be spot on

A lot of water

HD

[#10]

Don't know if this works, but:-

It was commonplace in Malaysia, many years ago, to collect water from a dripping tap on a 24 hour basis.

"Showers" consisted of pouring a small bucket of cold water over you.

Apparently, water was very expensive (and metered), but this method defeated the water meter.

[#11]

Thanks for the replies folks.

@Baldrick1 

First I would try to get this witnessed, preferably by Severn Trent before they remove the meter

That's my plan, although sods law it's behaved itself so far today. It's not so sunny today so I'm fairly convinced that has something to do with it. You have some interesting thoughts on pulse counters though.

This could be disproved if, when you were seeing the meter clocking up at a silly rate, you turned a tap on, which  would move the paddle wheel and the rate would return to something more sane.

I'll try that next time it happens.

The problem with pulse counters is that they are bi directional. If this is the case then any stray pulses getting in to the electronics will get counted. However in real world terms these pulses will be at a slow rate,

Im not sure the meter operates at that slow a pulse rate. The meter display has 5 digits after the m3 dec point, so gives a 1/100th litre resolution. Say refilling a toilet cysten is 10l in 20 secs, thats 500 pulses/sec under normal circumstances. The sort of flow meters I'm familiar with in process automation would be operating in the 10s of hertz. In that case filter circuits can easliy take out any electrical noise. If its normally operating at 500 hz then filtering is going to be more difficult.

edit: ignore that , my calcualtion is wrong ,its only 50 pulses/sec, so it ought to be possible to isolate any electrical noise with a filter but maybe as its a consumer item they havent bothered with any filtering...

[#12]

Expanding on that thought, I suppose that if the paddle wheel happened to stop at the trigger point and there was no dead band between the pulse generator on and off trigger points, possibly due to the transmitting magnet losing its strength over time, the pulse trigger might oscillate and output a string of pulses. This could be disproved if, when you were seeing the meter clocking up at a silly rate, you turned a tap on, which  would move the paddle wheel and the rate would return to something more sane.

@Baldrick1 

You might be on to something here. I happened to catch it showing flow when no taps on, a quick slight open/close of the kitchen tap stopped it.

Its overdue for replacement according to ST so I presume its battery is run down.

[#13]

@MisterW "edit: ignore that , my calcualtion is wrong ,its only 50 pulses/sec, so it ought to be possible to isolate any electrical noise with a filter but maybe as its a consumer item they havent bothered with any filtering..."

How strange, my electricity supply happens to be exactly 50Hz...

[#14]

Good point

[#15]

@MisterW Depending on the size of the expected bill I would not be having the meter simply changed. As for arguments, presumably you've read this - towards the bottom. If you do suspect the meter is reading wrong, get it tested but be certain about it, £70 is £70! Edit. Forgot to add the VAT!

For the future, you could always do what I did years ago and fit another water meter in line with the company one (I use to have a very large koi pond). £40 or so nowadays, if you can fit it yourself.