Elektrometal EM-eG4 - how to read gas consumption remotely, integration with Home Assistant

hello, the gentlemen from pgnig came and replaced my new meter (one year old) with a new one - model from the manufacturer Elektrometal S.A. type EM-eG4. In the previous one, I had a pulser (reed switch) fitted and the pulses were sent via SonOff (I used a door-opening sensor) to my Home Assistance. Well, and a flop, because now the meter is different and the pulser (Metrix NI-3) does not fit. The new one has a communication module and supposedly communicates with the mobile network once a day.

Do you have any idea how to remotely read the gas consumption from this meter? The documentation says it has different types of communication (min LoRa), but I don't know if it's standard and if it's even possible to communicate with it.

https://elektrometal.eu/products/1318

https://elektrometal.eu/catalogs/all/plgazownictwo.pdf

The fitter said they are replacing en masse with these particular meters (the debt collectors will go on the pavement...).

+1... same story with me

I would like to make a comment to this thread. The company seems to be replacing gas meters en masse. Same problem at my place, small integration with HA via Sonoff. Everything was laughing beautifully until yesterday. After reading the documentation, very vague, I'm going to try to download the readings via CC1101 868 MHz, ESP32 and ESPHome software, ESPHome custom components by Szczepan. Maybe someone has a better idea or has already tried something.

>>21752558 If it were possible to intercept the radio signal, it would be cool - I do that with the water meter I have at home. But this meter is the worst kind of crap - it has neither a magnetic nor light pulse generator, nor does it communicate by radio. It only has a SIM card to communicate with the GSM operator.

The only solution, in my opinion, now is to mount some sort of battery-powered ESP module inside this box next to the house, which would read the meter status with the camera. But this is against the regulations - I found out from the gas supplier, anything that is mounted there: 1) has to be EX approved (or something like that); 2) has to be approved by the gas company.

If I could buy a reasonably priced meter of some sort that I would fit to my installation at home (with the help of some professional, of course), that would be lux.

I found information that the gas meters are equipped with an optical interface in IEC 62056-21 standard. This is an optical serial port.
There is a component for ESPHome that can be implemented on the ESP8266 or ESP32 board.
Now only the optical head in ATEX design and here I found a company ELGAS, model HIE-03, production from Czech Republic. But for now I am waiting for a reply from PSG. There is also the EU directive 2023/1162, which says that the end customer has the right to read the readings from the measuring device. PSG is replacing gas meters with KPO and therefore EU money. I will wait for an answer before taking steps on what to buy. I would also advise you to do so and ask the gas company for permission.

This is very interesting... I called the manufacturer about this gas meter and the gentleman from technical support firmly answered me: 'This gas meter has nothing, only a GSM module'.

It would be great if it turned out that there was something there after all - even if I had to buy a module.

The gas company agreed - I had already received permission from them to install a module for the magnetic pulser, and I had already made preliminary enquiries here too. There is one condition: ATEX.

I would be grateful for any information on what has been established.

jacekiwanski10 wrote:

I found information that gas meters are equipped with an optical interface in IEC 62056-21 standard.This is an optical serial port.There is a component for ESPHome that can be implemented on the ESP8266 or ESP32 board.

It is just that in gas meters in Poland the OPTOGAZ optical connector of the IEC 62056-21 standard works with the Smart-Gas protocol (ST-IGG-0201:2023) and not DLMS/COSEM.

The protocol is secondary, finding an ATEX optical head is probably impossible, SICK used to produce one but it is now discontinued. With the NI 3 reed switch it was straightforward because it met the requirements, and the Sonoff box was mounted outside the explosion zone and that was it. The head is an electrical device and must be EX. I got a laconic reply from PSG that it was planned to have access to read the eGas meters. Just not sure when, so I asked the question if it would be via wM-Bus or via API, a question I didn't get an answer to because they don't know.

The new meter’s pulser won’t work. Remote reading likely requires the utility’s platform or a compatible gateway, since it communicates via LoRa or mobile network. Best to check with Elektrometal or your gas provider.

I found information that the gas meters are equipped with an optical interface in IEC 62056-21 standard. This is an optical serial port.
There is a component for ESPHome that can be implemented into the ESP8266 or ESP32 board.
Now only the optical head in ATEX design and here I found a company ELGAS, model HIE-03, production from Czech Republic. But for now I am waiting for a reply from PSG. There is also the EU directive 2023/1162, which says that the end customer has the right to read the readings from the measuring device. PSG is replacing gas meters with KPO and therefore EU money. I will wait for an answer before taking steps on what to buy. I would also advise you to do so and ask the gas company for permission.

jacekiwanski10 wrote:

the end customer has the right to read the readings from the metering device

And the new gas meter does not have numbers on the display?

I am after a conversation with a nice PSG employee who informed me that all communication ports on the new gas meters are blocked to third parties. And it is planned to make the data from the eGas meters available, only for now it is not clear when and in what form. I have come up with another way of monitoring gas consumption. When I test it, I will describe it

It was so much fun with the magnetic pulse based meter that I had a real time measurement with an accuracy of 0.001 m^3. This is not going to come back with the crap they have installed now. Reading via API (even if they make it available), or via the website (some web scraping) is definitely not going to be the same anymore.

Added after 2 [minutes]:

I'm wondering if I should buy an old style bellows gas meter (about £300) and with the help of some installer if I don't get myself one already in the house.... Maybe it doesn't make sense, but I don't see any other option. Measuring gas consumption was very important to me....

It seems that the easiest way would be with the ESP32-CAM. Only the housing needs to be made
https://github.com/jomjol/AI-on-the-edge-device

I have the same problem. I have spoken to the manufacturer Elektrometal and also confirm that there is no communication of any kind with the meter. I found out that full encryption is used at all stages of transmission and the local interfaces are deactivated by default. Communication via the IR port is also subject to full encryption. For security reasons, the PSG does not share this data, and the port itself is not designed for integration into data acquisition systems - it is purely a fail-safe service function. The only available method of data acquisition remains the physical reading of readings from the LCD display.

On the eGas Meter project page, PSG has a description how to correctly read the status of a smart gas meter . There are three meters listed:
* Elektrometal EM-eGx
* Apator Metrix iSMART2
* Intergas IGAZ BK-G4

I'm thinking if maybe either of the latter two has some kind of open, additional method of communication. Maybe some of you have one and have looked into the subject? Well, unless it's a PSG requirement that there's no way to communicate with the meter, in which case it probably wouldn't be up to the manufacturer. I'm just thinking to myself, if there was any meter installed by PSG that gives the ability to communicate, would it be possible to request a replacement meter, even for a fee.

-- Update 9.02.2026 --
I'll add right away, because I've contacted all these companies - each responds that the meters are blocked for consumer integration.

On the Apator Metrix gas meter above, I received some additional information from the manufacturer:

* No pulse output: The iSMART2 model is a smart gas meter in which the traditional mechanical counter with magnet (allowing the use of reed sensors) has been replaced by an electronic measuring module. This design does not provide for the installation of external pulse transmitters.

* Closed communication system: The iSMART2 gas meters are devices operating in a closed telemetry system based on the DLMS protocol with data transmission over the NB-IoT network. This system uses dedicated APNs defined by the Ordering Party, i.e. the Polish Gas Company, and is incompatible with previous generation devices (mechanical gas meters, e.g. with the Unismart1 module, which used an open radio protocol in accordance with the EN1375-3 standard, in the 868 MHz band).

* Lack of consumer interfaces: the gas meter does not have pulse diodes (such as in electricity meters), open infrared (IR) ports or Bluetooth/Wi-Fi modules that would allow the device to be paired with home automation systems.

The management of the reading process, the administration of the ICT infrastructure and the provision of gas consumption data is the sole responsibility of the Distribution System Operator, i.e. the Polish Gas Company, and in accordance with the applicable regulations and data security policy, only authorised systems of the Operator have authorised access to the remote reading of these devices.

I came across this thread because I received a notice yesterday that my gas meter would be replaced. After contacting the installer, it turned out that it was the electronic version that would be fitted - however, the installer refused to tell me what specific model of meter it was to be. He referred me to the PSG.
And I have been monitoring my meter with a reed switch for many years and would not like to give up these measurements.
My question is - is the purchase and installation of "my" reed switch bellows meter the only sensible alternative.
I've done a bit of searching and, all in all, you can't see the possibility of buying an unlocked electronic gas meter - leaving aside, of course, the expected cost absurdity of buying such a device Perhaps the oblique-eyed friends would help with their inventions on Ali?

Unfortunately on a reed switch is the only option. Remember, it's gas - everything has to be in EX, because if something goes wrong, the insurer will bail out. I went the other way. I made an energy monitor from Shelly 1PM that converts this to the gas consumption of my gas boiler. 99.4% accuracy.

jacekiwanski10 wrote:

I went the other way with Shelly 1PM I made an energy consumption monitor that converts this to gas consumption for my Gas boiler. Accuracy 99.4%.

How did you do this? At my home I have a Home Assistant and Viessmann ViCare integration. It gives the modulation of the burner. From this I calculate the current gas intake and consumption over time. But it seems non-linear, plus this integration has a delay of about 2 min. Therefore, the effect is meagre in terms of accuracy. It is not completely out of touch with reality, but too far from the truth.

I couldn't use the integration from Vaillant because I have an old boiler without an ebus, so I opted for Shelly, in which I uploaded a script and ran MQTT, plus a few entries there and then in the HA, and Shelly sends to the HA at 10 s intervals the amount of m³ consumed with a resolution of 0,000 m³. The accuracy with me is 99.4%, and the meter intentionally overestimates the results by this 0.6%. I correct it from time to time and am very satisfied. No unnecessary cans of farts etc.

Then I understand correctly that the furnace provides some kind of pulse output that counts the gas consumption?
I'm referring to what this circuit counts.

Well no, no pulses, Shelly between loads in watts and converts to m³, no interference with the boiler, it took me a while to calibrate the system, but since 24 December it has been laughing. And to date the adjustment has been about 6 m³ of gas for a consumption of 950 m³. This winter has taken its toll and this was nicely visible in the consumption graphs, in addition to the consumption Shelly provides a "Burner operation status" entity. With it, I could see how often the boiler turns on and how long it takes for it to heat up cleanly.

Very clever. I hadn't come across that, and I also have a Shelly PM1 plugged in before the boiler. Just how did you come up with the relationship between load and gas consumption? There's a circulating pump in the boiler too, for example, and probably various other controls (relays, some external sensors). They can randomly affect the load.

These are my examples, emphasis mine: 5.4 W - standby, 91-95.6 - operation of pump alone, 96 to 105 - operation of burner and pump. And here the more, the higher the modulation. If you have a Shelly plugged in, observe for yourself how it is for you. The more observations and conclusions you make, the more accurate your measurements will be

Thanks for the inspiration. I probably wouldn't have come across that. I have graphs of modulation and power consumption; then I can try to infer something. And you have an accuracy that I don't think even my pulse counter was that precise. It was always losing 2-3 meters per month (1 m³ is 100 pulses). You also have to take into account HA updates etc that break contact with the ESP.

In my case, all the metering logic is in Shelly and even if I disconnect the HA for a whole day, I don't lose consumption - it will pull the current value. It is Shelly that is the meter independent of the others, and in the HA there is only the device class set.

I don't know how you came up with this, but I bow my head. In my Viessmann, you can also see the dependence of the power consumption on the modulation of the burner (yellow is modulation, green is the power consumption of the furnace with all the accessories connected to it: pumps, valves, etc.). The modulation here is provided by the ViCare integration and is time-delayed (changes updated probably every 2 minutes).

We've actually started a very loosely related thread already, and maybe there should be a new one, but I'm wondering what electrical component in this cooker is controlled in parallel with the burner, that the gas intake can be calculated based on it. Some sort of fan? My point is that if it were isolated and only its power consumption was measured, it would be easier to estimate or adjust the power consumption conversion rate to the gas consumption rate per unit time as a starting point for calculating gas consumption. I'm assuming this is where some numerical integration (in HA) comes into play, unless you've somehow made it simpler .

Let me give you my observations. Well, the boiler, together with the radiators, is a sort of fairly fixed system. From it, you can measure how much gas the system consumes; now you just need to change how long it did it, and the time-flow entity is ready. I'm not imposing on you how to measure the burner time, the important thing is to measure it as accurately as possible. From my observations, this arrangement is quite stable. I would also add that I created a helper - a utility counter - based on my Shelly entity and only added it to the Energy panel; the helper is easier to calibrate in the development tools. Your focus at the start should be to read as precisely as possible when the burner is running, not with what modulation.

>>21851509 Am I understanding you correctly then, that the burner modulation is already secondary and doesn't have much impact on the accuracy of the measurement? I'm looking in the boiler manual for where to plug in. At the moment I don't know anything for sure yet, but the "universal gas regulator" connection is shown there. If by chance it is a voltage regulator in some range, maybe that would be such a surest source already.

If I can just give you a hint, you are in a better position than I am. My 2001 boiler has few studies on it, so I came up with something of my own. You're better off, if you're bothered by delays, they're caused by the cloud. There is a design for Viessmann for local readings via OT and your boiler should support it. Maybe you'd better focus on that, you'll have more information. There is also the Open3e project for your internal bus. And as for your question, the modulation in the boiler matches the heating system and is fairly constant. I don't know how you control the boiler, is it a temperature setting for example in the living room or a heating curve? If a heating curve, then here I would focus on the burner time and the extracellular temperature.

jacekiwanski10 wrote:

And to date the adjustment has been about 6 m³ of gas for a consumption of 950 m³.

Shouldn't you by any chance be basing it on the kW calorific energy of the gas which is variable instead of m3? The higher the calorific value, the shorter the cooker will run to heat the same amount of water. At least in theory