New dataset on EMI – Data on metering configurations

  • Last post 12 January 2023
Matthew Keir posted this 02 October 2016

Agents, third parties, and new entrant retailers have asked for information about metering configurations at ICPs throughout the country. These parties are investing in systems and processes, and it’s helpful for them to know all of the options up front rather than discovering each unique configuration one-by-one as they engage with consumers or use the ICP connection data API.

We’ve released a file detailing unique metering configurations at ICPs and the number of ICPs with each configuration. This information is provided at meter register level for any active meter in the registry at any installation at active ICPs. The register information is provided as a space delimited list with energy flow direction, register content code, and period of availability. More information can be found in the registry functional specification.

The file will be updated and released with other retail reports each month. As with all files in the datasets folders, they are also accessible via our FTP site.

  • Liked by
  • jofajafa
  • JessielGonzalez
Order by: Standard | Newest | Votes
msouness posted this 01 February 2017

Is there any good reason for having 1580 I flow registers allocated against UN / D / N etc content codes?

  • Liked by
  • jofajafa
msouness posted this 19 February 2017

I guess my main issue is with PV registers that may typically be allocated PV1 profile having UN24 register codes in place of EG24.

On another note, interesting to see a number of zero hour registers.  Suspect these are three phase meters installed in place of two single phase meters.

Following on from this, is there a report that indicates the number of whole meters not used for submission?

Matthew Keir posted this 24 February 2017

No - after looking into this, there is no good reason. This is an MEP compliance issue - I'll pass it to the right people internally.

For injection meter registers can only be:

  • I-7304-24 (with an HHR profile code populated by the trader), or;
  • I-EG-24 (with a PV1 or EG1 profile code populated by the trader)

Zero is not a valid period of availability. A UN register content code must always be 24. If the register is programmed or exists it should show in the registry. However, observing this in the registry data would not tell you if that channel was connected to any actual load (or any injection source), to see this you'd need to look at the consumption data for the ICP.

There is no report that lists the numbers of separate installations or separate meters installed behind ICPs that are held in the registry. We thought the most value was captured by reporting register details of all installations and meters at each ICP - and how common these arrangements are. Other meters at ICPs that are not used for reconciliation won't have details held in the registry.

If you can explain a bit more about the value in this we would consider this.

  • Liked by
  • msouness
msouness posted this 26 February 2017

For these combinations I have to scratch my head a bit:

X-EG-24 (generation, but not into the network???)
X-UN-18 (uncontrolled, but not available for more than 18 h per day???)
X-N-24 (a really long night?)
X-N-20 (night lasts 20 h per day?)
X-NC-24 (night register, fully controlled, but not controlled?)
X-CN-24 (controlled but available 24h per day?)

To a lesser extent, the following appear to be a bit odd:

X-SENW-10 (10 hours weekdays, 24 h weekend)
X-SEOP-18 (7.5 h per weekday)
X-SEOP-24 (7.5 h per weekday)
X-SEPK-24 (6.5 h per weekday)
X-SWDD-24 (weekday day)
X-SWDN-24 (weekday night)
X-SWDPK-24 (8 h per weekday)
X-SWED-24 (day weekend)
X-SWEN-24 (night weekend)
X-WD-8 (day is 8 hours only?)
X-WDOP-24 (off peak is 24 hours a day?)
X-WE-8 (the weekend only has 8 hours per day?)
X-WWDD-24 (winter weekday day is 24 h - what about night?)
X-WWDN-24 (winter weekday night is 24 h - what about day?)
X-WWDPK-24 (peak period is all day?)
X-WWED-24 (winter weekend day is 24h? - what about night?)
X-WWEN-24 (winter weekend night is 24h? - what about day?)
X-NC-20 (can night be 20h?)


  • Liked by
  • jofajafa
Ron Beatty posted this 06 April 2017

Apologies for the delay in replying to this post. Most of the combinations listed above are incorrect, we are investigating these along with a few others currently.

  • Liked by
  • msouness
  • jofajafa
msouness posted this 13 April 2017

Regarding whole meters not used for submission, my interest is to find out how many meters are active, but submit no volume due to discontinued channels. In particular meters that have no connection to customer load.

Ron Beatty posted this 18 April 2017

The registry contains metering information to the register level, however, market systems currently only require aggregated electricity volume information.

The only participants that would be able to determine "how many meters are active, but submit no volume due to discontinued channels", is the metering equipment provider and the retailer/purchaser.

msouness posted this 18 June 2017

I'm interested in finding out how many meters are deployed that are redundant due to legacy channels no longer being used.

  • Liked by
  • jofajafa
jofajafa posted this 19 September 2017

Shouldn't there be business rules to prevent creation in the registry of meaningless content codes such as CN-24 or N-24? There's a great deal of unnecessary complexity in wading through so many different configurations when some combinations are invalid.

  • Liked by
  • msouness
  • Clive Gifford
Nicole Gagnon posted this 19 September 2017

We are currently consulting on proposed changes to the register content codes and how they are updated/used in the registry. All submissions are welcome, the closing date is Friday 29 September. Details are on our consultation page.

  • Liked by
  • msouness
  • jofajafa
jofajafa posted this 29 September 2017

Thanks Nicole, we have submitted. I've taken the Metering Configurations file and reworked it to provide some insights.

msouness posted this 08 October 2017

It’s certainly time for some consolidation of register content codes and hours of availability. How many networks cut different tariff rates for different hours of availability for controlled load registers?

  • Liked by
  • jofajafa
Clive Gifford posted this 25 October 2017

Can one of the Authority staff please explain why there is no October update of the metering configurations file published yet.

Has this been delayed or postponed, perhaps wishing to avoid the effects of "garbage in, garbage out" as alluded to in the submission I made concerning register content codes, including thousands of non-existent meters and their associated register content codes, at least on TLC's network?

As of today, 26 October 2017, the last update was on 15 September. I understood an update was published on the 15th of each month.


  • Liked by
  • jofajafa
Matthew Keir posted this 25 October 2017

Hi Clive,

Thank you for making us aware of this. The September files are now available.

We had a small issue occur in our data warehouse that happened to coincide with the automated processing that runs after the 10th business day each month. We hadn’t realised this had impacted any published files. This issue impacted all the monthly files published in retail datasets, not only the metering configurations file.

One of the reasons EMI exists is to support users to provide informed consultation responses. It’s great to hear that you’re making use of this information in this way.

To be clear, the files are published after the 10th business day - this is not always the 15th of each month.




  • Liked by
  • jofajafa
  • Clive Gifford
msouness posted this 28 February 2018

Hi Matt,

Is there a new location for the meter configuration dataset?

  • Liked by
  • jofajafa
msouness posted this 28 February 2018

- found it


Matthew Keir posted this 28 February 2018

Sorry - the heading I used may mean you missed it. See point two in the "Changes to datasets on EMI – NSP table, metering configurations, market structure and shape files" discussion.

msouness posted this 24 August 2018

Interesting to see there are

138 residential ICPs with "X" flow EG meter registers

107 residential ICPs with "I" flow UN meter registers



msouness posted this 11 October 2018

It may pay for some of the MEP's to take a look through their serial numbers to identify data type conversion to exponents due to excel's treatment of large integers.  

2.61E+11 (IN), 2.62E+11 (IN)

Matthew Keir posted this 11 October 2018

Hi Malcolm,

Thanks for the note.

Just to be clear where this conversion issue is occurring - are you suggesting MEPs are populating the registry with serial numbers after opening in excel, and that what is causing this issue?

We're still pushing for improvements to reduce incompatible meter configuration information.





  • Liked by
  • msouness
msouness posted this 23 October 2018

Just been working through some serial number missmatches - found some Exponents. One MEP had three exponents as serial numbers, but are now fixed in registry.

msouness posted this 19 June 2019

Anyone have a good reason for the content code EG20 to be in Registry?  

jofajafa posted this 20 June 2019

"Anyone have a good reason for the content code EG20 to be in Registry?"

It's disappointing that not only has the morass of legacy CC/POA combinations - many combinations of which as Malcolm has demonstrated are nonsensical - not been tidied up, but there apparently are no validation rules to limit the creation of new redundant combinations.

  • Liked by
  • msouness
msouness posted this 01 October 2019

Great to see significant improvement of register content codes:

  56 I Flow non- EG (or 730x)

  374 I flow EG with less than 24h

  148 X flow EG24

  70 X flow non uncontrolled with 24h

  161 0h registers

However IN24 is still substantial at 316,902 registers.

  • Liked by
  • jofajafa
msouness posted this 28 October 2019

Relocated meter register conversation to new discussion


  • Liked by
  • Matthew Keir
  • jofajafa
jofajafa posted this 11 January 2023

Hello EA team. There doesn't appear to have been any recent updates to this data; the file available to download is from April 2022.

This is disappointing as there are frequent additions to metering configurations (for example, I've noticed a lot of DIN and NIN new register types added in support of EV charging) that are difficult to track proactively. With a file like this, I have some ability to ensure pricing to meter configurations based on the registers in market. Without it, we often have to take a reactive approach to finding new registers on prospective customer sites.

Is there the potential for an updated file shortly?


  • Liked by
  • msouness
Phil Bishop posted this 12 January 2023

Too much work to be done and not enough people to do it. There's about a week's work (for one person) required to rebuild the processing and publishing of this dataset. It was stopped when we finally pulled the plug on our old SQL Server instance almost a year ago. It remains on the TODO list to be restored. I will try to get it bumped up the priority ordering but can't make any promises as to when it will happen.



jofajafa posted this 12 January 2023

Thanks Phil. I appreciate there is some work involved.

I would much prefer not to be making use of this data set but unfortunately the combinations of content codes and periods of availability present in the market keep proliferating. Within recent years the number of combinations was in the 80s, now there's well in excess of 100. Many of them used on just a few ICPs when an existing combination would be appropriate, so the optimal situation would be that there were principles and controls on what is implemented. Also are still some nonsense combinations being implemented, such as period of availability being zero, or 24 hours for a controlled register.

  • Liked by
  • msouness