Customer convinced that generation meter is not calibrated

[ktech]

Hi,
I have a customer that is convinced that his generation meter is not calibrated and possibly not Ofgem approved. The type of meter we most often install for domestics is a L&G E110 5235 (A in this case). I quit happy the meter is approved and calibrated. For the customers sake we replaced the meter and notified of a meter change with the FiT provider but now the customer says this new one is even less accurate.

To cut a long story short they are using a clip on meter to measure against the generation meter. We provided an OWL device that they are using to compare and also they have tried a standard monitor from British Gas and the likes. They customer reports that both the clip on meters show the same but this does not correlate with the generation meter.

I know the clip on meters are not as accurate as the generation meter and have sent the customer all the supporting information from Landis of the various test approvals and Ofgem approval for the generation meter.

Do any instances exist when the clip on meters are particularly out, maybe with a high proportion of non-resistive loads etc..? Just remembed they customer is on storage heaters if that helps.

 

[JulianC]

Don't the clip on devices assume a fixed mains voltage of 240v or 230v and hence only provide an approximation of consumption. I know the OWL allows you to select either 230 or 240v for mains voltage but none of them accurately measure the actual voltage.

 

[yorkshiremike]

Have you seen which cable he is clipping on to? - I'm thinking something daft like he's clipping onto the grid? Other than that, all the clip on monitors we've tried have been out by 5-15% (typically under reading at low PV outputs and over reading at high outputs). I've heard of an inverter apparently clocking up 100W+ with a clip on monitor at night, apparently due to the power factor when 'sleeping', but don't know if this is true or an urban myth.

 

[jason121]

I agree with Julian C voltage needs to recorder as well.

 

[whinmoor]

There is a list of approved generation meters somewhere...

The important question is how his system is performing against SAP predictions. Mine was 36% up for the month of December so I'm pretty pleased.

 

[JD6400]

Not sure about other inverters as we have only ever supplied Fronius , but can you not back the meter reading up with the inverter log ?
At least that will be a bit more accurate than a £30 clip on meter .

 

[ktech]

I can't see any issues with generation and sap and the customer hasn't raised any but is convinced the gen meter is not calibrated as it does not tally with the clip on meters.

I guess I might suggest they have a search for calibration certificates for the clip on meters! However that said we supplied the Owl and in general customers have given positive feedback on these and appreciate that they are indicative tools and not 100%. The Network Owl registration process does allow you to enter a voltage value (web interface) however i'm not convinced this actually does anything - will email them to check.

I think this is probably a case of the customer latching on to the fact two clip on meters read the same output and this differs from the gen meter so the generation meter must be at fault.

 

[ktech]

I did also explain to the customer that the FiT provider require the meter make and model of the generation meter and this will be checked by them for been on the approved list.

 

[jason121]

Get the customer to call the meter company and let them explain that the OFGEM meter is more accurate.

 

[ktech]

I think probably one of the best methods (keeping it economical) of monitoring is using the pulse from the meter as at least the two will be in check. Pitty Owl don't offer this as the web data / interface on the network range is very good.

 

[JulianC]

All the clamp on devices do is measure current - probably not with much accuracy and almost certainly not calibrated. In order to convert the current measured into a generation figure it obviously has to be multiplied by voltage and time. The voltage is not measured, it is assumed at some fixed figure. Obviously in reality the voltage is bouncing up and down like a yoyo. The generation meter will be measuring both the voltage and current and will of course be more accurate. Moreover, it will be calibrated , ofgem approved blah blah blah...

Presumably the gen meter is reading less than his clamp on kits, otherwise he'd wouldn't be complaining. Customers....bless 'em.

 

[ktech]

Yes i understand the clip meters are reading less. I've given the customer the link to this thread so they can get some reassurance and will query with Owl as well.

 

[Marvo]

Doesn't the Landis come with a calibration certificate? If so I would just inform the customer he is wrong in his assumption and leave it at that. Sounds to me it's one of those scenarios where you'll never convince him otherwise so I wouldn't even try.

 

[ktech]

Interestingly never received a cert ever when purchased (purchased through Segen). Although I don't see this as a problem as they are certificated so one can be downloaded.

 

[Eco Deal]

never had a cert with any gen meter bought!

 

[liquidity]

What is the inverter saying has been produced?

 

[Gavin A]

Hi,
I have a customer that is convinced that his generation meter is not calibrated and possibly not Ofgem approved. The type of meter we most often install for domestics is a L&G E110 5235 (A in this case). I quit happy the meter is approved and calibrated. For the customers sake we replaced the meter and notified of a meter change with the FiT provider but now the customer says this new one is even less accurate.

To cut a long story short they are using a clip on meter to measure against the generation meter. We provided an OWL device that they are using to compare and also they have tried a standard monitor from British Gas and the likes. They customer reports that both the clip on meters show the same but this does not correlate with the generation meter.

I know the clip on meters are not as accurate as the generation meter and have sent the customer all the supporting information from Landis of the various test approvals and Ofgem approval for the generation meter.

Do any instances exist when the clip on meters are particularly out, maybe with a high proportion of non-resistive loads etc..? Just remembed they customer is on storage heaters if that helps.

One device is accurate and calibrated, the other provides a rough approximation based on a guess at what the grid voltage reading might be.

I'm not really understanding why you're having trouble explaining this one tbh, but try this from Owl themselves

Solar PV installations operate outside the bounds of the electricity monitor parameters for most of the day. At very low power levels (below 500W) the monitor accuracy will be significantly impaired.

Accuracy​	<1A​	1A to 3A​	3A to 71A​	>71A​
Current RMS​	Not Specified​	Better than 10%​	Better than 5%​

and 2 different devices operating on the exact same principles will likely give roughly the same readings, but both readings will still be wrong for the exact same reasons as each other - ie if the voltage is 250V and the meter is assuming 240V etc. and bear in mind that when the inverter is outputing at full power then this will also be pushing the voltage up and increasing the error in the devices calculations.

 

[FB.]

As an earlier poster said: does the inverter agree with the generation meter? Mine are over 99% in agreement; I have a L&G E110 and an Aurora. My incoming electricity meter belonging to the utility company is the same L&G E110. I recently mentioned that I soon planned to stop contributing to PVOutput.org because the data being uploaded by most people's devices was often unbelievable compared to the few people who take readings from their generation meter. The Sheffield Solar Farm insist on readings only from the generation meter.

 

[ktech]

I think once you get into looking at invertors and generation meter readings it becomes another can of worms, i.e. is the inverter reading relating to the DC or AC side, voltdrop issues as well as reactive loads etc..

 

[JD6400]

I think once you get into looking at inverters and generation meter readings it becomes another can of worms, i.e. is the inverter reading relating to the DC or AC side, voltdrop issues as well as reactive loads etc..

Please don't take offence , but how can you fit the gear with out knowing how to use it / set it up .
What inverter is it ?
My personal system is only about 3 units out between the both of them after over a year .

 

[ktech]

Please don't take offence , but how can you fit the gear with out knowing how to use it / set it up .
What inverter is it ?

Eh? Sorry you have completely lost me...

 

[JD6400]

I think once you get into looking at inverters and generation meter readings it becomes another can of worms, i.e. is the inverter reading relating to the DC or AC side, voltdrop issues as well as reactive loads etc..

I am probably wrong , but this post sort of seems to suggest that you have a problem in reading the information from the inverter ?
As i have already said , i am not sure about other inverters but on fronius ones it gives a very easy to understand read out .
Again sorry if i have it wrong , it was not meant to offend .

 

[ktech]

Hi, sorry but you are indeed wrong, no not a problem reading off the inverter. As per original and previous posts issue is with the difference between generation meter and clip on meter monitoring devices, invertor readings never mentioned or implied.

 

[JD6400]

Fair enough , i have had a few tonight .
But please do bear in mind that if the output from inverter matches the generation meter readings , that alone should pacify the customer .

 

[SolarCity]

Although bear in mind that inverter displays can be out too. The only reliable measure of how a system is doing is by looking at the generation meter.

 

[Worcester]

We have 5 different monitoring devices on our various PV generators here.
The Inverters (x3), meters (3 x generation, 2 x import, 1 x export) , immersun and calibrated Solar PV Monitoring System | OpenEnergyMonitor (6 x CT montiors) all tally within about 1%
The two energy monitors, Scottish Power and current Cost have an error factor of between 25% and 50% based on ACTUAL production / consumption.

Speak to them (SP and CC) and they will tell you that they aren't accurate and onlu designed to give the user a guide / base line / indication. The CC version with the pulse meter attachemnt is more accurate.

I VERY much doubt it is the meters at all

 

[FB.]

Although bear in mind that inverter displays can be out too. The only reliable measure of how a system is doing is by looking at the generation meter.

My inverter reads about 0.5% higher than the generation meter. This small difference is probably a result of losses in the few metres of cable from the inverter to the generation meter.

 

[Gavin A]

My inverter reads about 0.5% higher than the generation meter. This small difference is probably a result of losses in the few metres of cable from the inverter to the generation meter.

there's a danger in applying your experience from one system across all inverters.

We were caught out with this assumption a while back, and it turns out that SMA inverters at least don't actually meter the outgoing AC electricity as such, they calculate what it ought to be in watts from the Ampage, and at least on their older models this conversion can be very rough and ready and not actually based on the actual measured grid voltage for some reason - I think it may even be that it's based on measurements of the DC side that are then estimated for the AC side based on some sort of efficiency factor, but I'm not 100% sure. Either way, apparently the SB series output data can be up to 20% out from the actual meter readings, though later models tend to be quite a bit more accurate.

The meter is calibrated, the inverter isn't, and neither are the vast majority of clamp meters, though there are some now that do actually also read the grid voltage and use that to calculate the actual wattage to within around 1% accuracy - Owl ain't one of them though.