Gas Meter Still Increasing Whilst Gas Isolated (Siemens/Landis+Gyr E6S)

Bendo

What are you waffling on about smart meters for Gerry when you have been told time and time again this isn't a smart meter.

samn9

Gerry1 said:

samn9 said:

Today I found error code 'b' for the first time. The meter seems to be getting worse and worse.

If smart meters were any good they'd phone home the moment the low battery code appeared so that the battery or meter could be replaced long before any problem occurred.

Out of interest, for newer smart meters, what is the impact of low/depleted battery? Have there been cases where it also starts overreading/incrementing erratically? Or do they usually just stop reading at all.

Something I find odd about my gas meter, is looking back on readings in 2023, there are a few months in a row where the gas reading did not change at all. I did not notice at the time as I wasn't reading my bills.

Gerry1

Bendo said:

What are you waffling on about smart meters for Gerry when you have been told time and time again this isn't a smart meter.

The point I'm making is that electronic gas meters (the ultrasonic ones as opposed to mechanically rotating ones) should give a far better warning of a low battery.

Mechanical gas meters with moving parts can't suddenly increment by large amounts.  Their  replacements should be at least as good and not have such a serious vulnerability.

Displaying a little b is meaningless: it should say REPLACE LOW BATTERY or something equally unambiguous.

If it's a smart meter then it should also phone home so that someone can be sent out to replace it long before it malfunctions.

If the new meters suddenly started seriously under recording they'd soon be withdrawn but, as MeteredOut and others have discovered, the industry has very little interest in fixing wonky meters that overcharge.

Grizzlebeard

Re low battery warning.

Would prob add to cost and complexity.

(Following thoughts are off the top of my head. I retired from electronics ~30 years ago before this gee-whizz tech had been developed. My brain-battery is also running on empty so I willingly concede to anyone who corrects me with better knowledge)

I assume they use Lithium or similar primary cells which maintain a fairly constant terminal voltage (@ low load) until it drops quite suddenly at end of life.

Actual end of life point is quite hard to predict because of manufacturing tolerances.

dv/dt detection on it's own miight be to late, so a low discharge backup standby might be required for the SOS call.

Or they could use Coulomb counting to squeeze a decent amount of E-juice out without false go/no-go detection.

Predicting end of battery life too soon is very expensive because it means sending out Engineers to change a lot of meters unnecessarily. Cheaper to allow the tank to run till it's totally empty and shug shoulders at consumer inconvenience....

(Edit--

P.S. for all I know the latest generation of meters might already be doing all this so it's "only" the millions of legacy meters that are still a problem.)

Gerry1

No reason why it should add to the cost and complexity.

Every cheap mobile phone and virtually every cheap portable DAB radio will have a battery status indicator, and it's usually far more meaningful than just a little 'b'.

In fact, the little 'b' on the gas meter suggests that the low battery detection is already present.

All that remains is for a low battery flag to be set and sent with all subsequent 30-minute readings.  That won't need an SOS battery, it'll just be the normal low powered ZigBee transmission to the comms hub next to the electricity meter.

Grizzlebeard

Gerry1 said:

No reason why it should add to the cost and complexity.

Every cheap mobile phone and virtually every cheap portable DAB radio will have a battery status indicator, and it's usually far more meaningful than just a little 'b'.

In fact, the little 'b' on the gas meter suggests that the low battery detection is already present.

All that remains is for a low battery flag to be set and sent with all subsequent 30-minute readings.  That won't need an SOS battery, it'll just be the normal low powered ZigBee transmission to the comms hub next to the electricity meter.

Thanks Gerry.

Stirred my thinking nicely. Cost/complexity noted and (largely) agreed. But not convinced battery status as currently done in mobiles/DAB allows 98-99% to be reliably squeezed from a Lithium Primary cell. I think most people have learned the hard way not let their cellphones drop below 20-30% on the fuel gauge.

I'm pretty sure the Network Ops don't want to send out Engineers to change Meters that might still have an extra 10-20% years of life in them.

For the the backup I actually envisaged a supercap(Vs) buck-boost maintained from the main battery(Vp) It's first purpose is not to supply the microamps needed for standby/data acquisition but to backup or maintain the milliamps for data transmission.

Equally importantly monitoring load sharing between (Vp) and (Vs) is a very reliable indication of when battery(Vp) is 99% discharged while still maintaining the (relatively)heavy current demand required for data transmission.

Also (almost) equally important - to your point - this only increases hardware costs by a fraction of a percent.

Maybe the hardware for this kind of approach is already hidden inside newer meters so a software update would do, but I doubt it.

Thanks for the stimulating conversation Gerry. (My 70++yo neurons are slow but still firing.)

To everyone else apologies for cluttering your reading experience.

Wibby.

MeteredOut

Gerry1 said:

No reason why it should add to the cost and complexity.

Every cheap mobile phone and virtually every cheap portable DAB radio will have a battery status indicator, and it's usually far more meaningful than just a little 'b'.

In fact, the little 'b' on the gas meter suggests that the low battery detection is already present.

All that remains is for a low battery flag to be set and sent with all subsequent 30-minute readings.  That won't need an SOS battery, it'll just be the normal low powered ZigBee transmission to the comms hub next to the electricity meter.

There's an assumption here that the b means low battery. It might well do, but I could find no details on what the different codes from the E6S mean (I also got E, C and d). ChatGPT gave some answers, but I think it was hallucinating.

I suspect b indicates the meter recognising something is wrong, and perhaps a low battery could be the root cause of that, but I doubt the meter itself has a way of measuring the actual battery level (although it might guess it based on voltage). 

Zandoni

The B flag does mean low battery and the meter was designed to have a battery replacement at 10 year intervals. Practically all the meters in the factory that incremented without passing gas were found to be poor solder joints on one the speed of sound sensors. I assume that as the battery as got older it has caused more poor joints to come to light. This is a fact I was there.

MeteredOut

Zandoni said:

The B flag does mean low battery and the meter was designed to have a battery replacement at 10 year intervals. Practically all the meters in the factory that incremented without passing gas were found to be poor solder joints on one the speed of sound sensors. I assume that as the battery as got older it has caused more poor joints to come to light. This is a fact I was there.

Do you if this is documented publicly, as I could find no record of it anywhere?

What do the C, E and d flags mean?

Gerry1

Zandoni said:

I assume that as the battery as got older it has caused more poor joints to come to light.

Also, perhaps the diurnal temperature cycling in a cold winter exacerbates the dry joint problem, condensation expanding as it freezes in much the same way that a cracked road surface becomes a pothole?