G83 pre install application

[JulianC]

If the output of the inverter cannot go above 16A, what business is it of the DNO if I have 3.68kWp or 36.8kWp of modules stuck on the end of it? So far all our systems have been under 3.68kWp anyway, so nothing has been flagged up by the DNO. However, tomorrow we fit a 3.9kWp system with an inverter limited to 15A output. It will be interesting to see if Central-Networks flag this up as an issue. I shall enjoy arguing the ---- with them until I am blue in the face.

Seems crystal clear to me that G83 allows up to 16A to be injected onto the grid. Surely if I fit an inverter with a max output of or below 16A there is no problem? Am I missing something blindingly obvious here that means, in these circumstances the size of the array has any baring on this whatsoever, regardless of what the inverter can actually deliver? OK, if I fit a bigger inverter, then potentially 16A can be exceeded, depending on the kWp rating of the array. However, 4kWp can be accomodated with an inverter limited to 16A no problem. Limiting the array to 3.68kWp only has a direct bearing on the amount of current injected onto the grid if the inverter is 100% efficient and grid voltage is 230v. ie, it ain't gonna happen.

Seems to me that the DNO's need to send their decision makers on a decent training course. Monkeys.

Cheerio

Dr. Ranty McRanterson

 

[JulianC]

Interestingly, Central-Networks seem to have changed the way they assess system size only recently. On a couple of our installs from a few months back, they simply multiplied the stated inverter max output current x 230v and confirmed this as the size of system when writing to the customer. On our last couple of jobs they are indeed taking the array size as the system size.

 

[TedM]

Here is the reply I have received from OFGEM on the questions posed back in November.
It has taken exactly two months to get this response.
I'm still waiting for any acknowledgement from ENA.

------------------------------------------------------------------------------------

Ted

In response to the rest of your email from November, we have outlined below our response.

1) refusal of connection or requiring payment to perform local reinforcement work

DNOs have a duty under the Act to connect on request and under the obligations of their distribution licence, licence condition 12 they have a requirement to offer terms for connection and use of system. In terms of charging for the cost of reinforcement work, Ofgem believes that it is reasonable for a customer to pay for the cost incurred of making a connection and therefore reinforcement work should be borne by the customer. However, we are unclear how that has a bearing on G83 where I believe the threshold sizes of generation are very small.

2) requirement to inform prior to installation and retain the right to refuse or impose other terms

It is again worth noting that DNOs have a requirement to offer terms for connection and use of system. The standard licence conditions can be accessed on our electronic public register on the Ofgem website. If you look for standard licence conditions in the open blue window of the e-register under the search box which is titled "Electronic Public Register". Under G83, there is a requirement for notification to the DNOs of an installation.

3) alignment with the ESQCR 2002

This is a question for the ENA who should be able to comment on the relationship between G83 and these regulations

4) compensation for DG

The DG Incentive is designed to encourage DNOs to connect DG efficiently. The DNOs are allowed to pass through 80% of the costs of reinforcement, i.e. get reimbursement for through UoS charges. This does not include any costs paid for by the customer but the incentive effectively covers the remaining 20%. The DG incentive only applies to DG that pays use of system charges – which we interpret to mean DG with an export meter. Therefore, not all installations that apply under G83 would earn a DG incentive. Finally, the amount that a DNO can earn under this incentive is capped against the amount that has been spent on reinforcement. Therefore, if a lot of DG connects which does not require reinforcement, the amount of DG incentive they earn is capped.

5) reports that Ofgem has taken steps in the past to prevent charging

We cannot comment on second hand information. We would need evidence or citation indicating our action in this regard.

Thereafter you included some specific points of clarification in the rest of your email. In general we feel these questions are best directed to the ENA as they directly relate to interpretation of the G83 standard, however we have provided whatever guidance we could below:

1. This should be directed to the ENA. However, we believe that there is some explanatory wording in the next standard up, G59 which may clarify this query.

2. According to the charging methodology, which is approved by Ofgem, customers must pay the cost for any required network reinforcement.

3. All of your query on what the 16A phase limit for G83 relates to should be directed to the ENA (though it may also be in the DG guide)

4. This should be clarified in the DG guide. In addition, in general, upfront fees are not allowed to be charged for the provision of a quote, however, we are not sure how this relates to G83 connections and therefore refer to the DG guide as a first point of information

5. We also feel that this may be in the DG guide and otherwise may be directed to the ENA.

6. As you are already in contact with someone from the ENA, it may be worth forwarding your query in its entirety and this response to a contact in the ENA and suggesting review of the questions and possible raising of this with DNOs at a suitable forum.

Kind Regards

Stacy

 

[TedM]

I'm still trying to understand some of the response but here is a bit more.

Under both items 1 and 2 a copy of the 'Standard Conditions of the Electricity Distribution Licence' referred to can be downloaded here:

http://epr.ofgem.gov.uk/document_fetch.php?documentid=15483

'Condition 12' starts on p47.

 

[TedM]

I received a reply this afternoon from the Energy Network Association to the questions raised about various G83 issues. Their reply is reproduced below with the original questions added back in in bold to aid readability. (I've not corrected any of the minor typos in the reply.)

Date: Tue, 22 Feb 2011 18:02:52 +0000
From: David Spillett <[email protected]>

Please find my response to your questions below.

Firstly In relation to the reference of refusal to connect above 1.9kW, G83/1-1 2008 caters for Small Scale Embedded Generators upto 16A per phase (3.7kW single phase and 11kW 3 phase). I am not aware of any DNO refusing connections above 1.9kW unless possibly there may be safety related concerns or equipment is not type tested as required by G83/1-1. G59/2 (2010) caters for connections above 16A per phase.

Q1 - DNOs are allowed to use their discretion to accept a >16A per phase system under G83 terms, based on their knowledge of the local network capacity. Once a DNO gives their agreement on this basis is this decision irrevocable or are they allowed to change the terms at a later date (when other changes may have happened in the meantime, such as additional G83 systems being connected in a locality) and then request that the original system complies with G59 - including requiring the customer to pay any associated DNO costs? Also clarify this matter where such multiple systems (such as wind and PV) may be connected at a single location over a period of time where, taken individually, each system comes below the G83 threshold but, taken together, they may exceed it.

DNOs were... and still are.. but the protection settings should conform to G59/2. Contracts can always be renegotiated - and if a customer doesn't like this, they can go to Ofgem for a determination.

However this is the formal background and I would not expect a DNO to change any terms retrospectively for domestic scale customers.

Where G83 applications in the same locality are genuinely independent, they would be treated as such and no reinforcement costs charged to any of the applicants. However, if "multiple systems...connected at the same location" means all within the same premises, it should be expect that DNOs will aggregate the electrical output of those individual sources of energy, in deciding whether or not the 16A per phase limit had been reached.

If multiple systems, each less than 16A are connected at different connection points but in the same locality then stage 2 of G83 is applied.


Q2 As regards multiple G83 connections, in a single locality at the same time, are the costs of any required network reinforcement, that may be required to accommodate these, to be passed on to customers or to be borne by the DNO?

The cost of any works to facilitate the connection/s will be assessed in accordance with the DNOs “Statement of Methodology and Charges for Connection to the distribution system.


Q3 Can we get clarity as to exactly what the 16A per phase limit for G83 actually relates to? Is it based on:
a) the Declared Net Capacity or the Total Installed Capacity of the system (as given on the MCS commissioning certificate) ?
b) the manufacturers quoted peak output of the PV panels (or other applicable prime mover 'nameplate' figure) ?
c) the maximum grid-tie inverter output?
d) if it is based on the inverter output then is it based on the manufacturer's maximum rating for the model of inverter or based on the (sometimes programmable) actual maximum inverter output figure
e) figures used at a - d above are usually given in Watts - so, in order to convert to Amps, is 230V or 240V to be assumed? I believe the definition in G83 specifically states "230V ac" but many DNOs seem unaware of this minor fact.
f) can you confirm the maximum Watts applicable to a 3-phase supply?

(a) For Inverter systems it is the output of the inverter so it would be possible to install a 3.6kW inverter with 4kW peak of PV panels under G83
(b) No see answer above
(c) Yes see answer above
(d) maximum continues rating
(e) 230 volts is the nominal voltage for the whole of Europe
(f) 230x3x16= 11,040 for practicable purposes this is 11kW, split evenly over three phases. The following is not allowed
phase 1 4kW
phase 2 4kW
phase 3 3kW
as two phases exceed 16A per phase

Unbalance is allowable up to to the following limit
phase 1 3.7kW
phase 2 3.7kW
phase 3 any value under 3.7kW


Q4 Question directed to Ofgem


Q5 How might the G83 threshold apply in a situation where a >16A per phase system has been designed to prevent no more than 16A per phase from being exported; e.g. a controller and dump load used for on-site heating for any excess that might be generated?

The system would need to be type tested for G83 to be applied.


Best Regards

David Spillett
Energy Networks Assocation
Dean Bradley House
Horseferry Road
London
SW1P 2AF

 

[edexlab]

Good work TedM thats answered a couple of outstanding questions I had ,that I've been asking central networks and sp manweb for 3 months and still no reply!

 

[TedM]

The next step will be to try to get all DNOs to recognise the stated standards so that we have a common approach for all of the UK. If any of you have any difficulties - especially with the G83 limit applying to inverter output - I suggest the first thing you try would be to quote this email.

 

[officerdibble]

What a load of -------s G83 is. It limits generation to 16A per phase and the allows manufacturers to build inverters that go up to 20A.
However, the kw output is derived from several factors, power factor being one and supply voltage being the other. It has been stated several times that 16A = 3.68kW, this is only correct within a pf range and essentially the supply voltage being at the nominal 230V, which we all know is not the case almost every where in the UK. The voltage tolerance in the UK from the nominal is +10% -6%, I believe at present. This puts the kW output from a 4kWp system between (-6%; 217V and +10% 253V) at 3.47kW and 4.05kW. The essential issue of G83 is that the generation is limited to 16A per phase. It is at the discretion of the DNO and they can allow systems up to 10kW under G83.

Be nice.

 

[Worcester]

TedM,

I really appreciate the work you've put into this to get these answers

Q3 Can we get clarity as to exactly what the 16A per phase limit for G83 actually relates to? Is it based on:
a) the Declared Net Capacity or the Total Installed Capacity of the system (as given on the MCS commissioning certificate) ?
b) the manufacturers quoted peak output of the PV panels (or other applicable prime mover 'nameplate' figure) ?
c) the maximum grid-tie inverter output?
d) if it is based on the inverter output then is it based on the manufacturer's maximum rating for the model of inverter or based on the (sometimes programmable) actual maximum inverter output figure
e) figures used at a - d above are usually given in Watts - so, in order to convert to Amps, is 230V or 240V to be assumed? I believe the definition in G83 specifically states "230V ac" but many DNOs seem unaware of this minor fact.
f) can you confirm the maximum Watts applicable to a 3-phase supply?

(a) For Inverter systems it is the output of the inverter so it would be possible to install a 3.6kW inverter with 4kW peak of PV panels under G83
(b) No see answer above
(c) Yes see answer above
(d) maximum continues rating
(e) 230 volts is the nominal voltage for the whole of Europe
(f) 230x3x16= 11,040 for practicable purposes this is 11kW, split evenly over three phases. The following is not allowed
phase 1 4kW
phase 2 4kW
phase 3 3kW
as two phases exceed 16A per phase

Unbalance is allowable up to to the following limit
phase 1 3.7kW
phase 2 3.7kW
phase 3 any value under 3.7kW

I am just quoting a customer for a roof (it's large) which faces just 10 degrees South of East, and he wants to get the maximum possible out of it.

Using PV*SOL, allowing for the orientation and roof slope and using local solar data, it reckons I can put nearly 4.8kWp (panel nominal output) on the roof , and still use a Diehl Ako 4300S inverter - the 4300S has a maximum continuous rating of 16.0A and 3,680W

Based on the reply you got above, do you agree that this should still meet the G83 requirement and therefore be covered by the usual post install process?


Update:
In this case since the output can be no more that 3.68kWp how does that fit in with FITS as the array size nominal is >4kWp, though it is limited by the inverter?


Thanks, Gordon

 

[SolarCity]

I had to go and check on the data for the Diehl 4300s inverter and you are right, it's maximum AC power IS 3680w (although this surprises me considering it can take 4.8kWp). Therefore it is acceptable to use this without prior G83 notification.

 

[TedM]

The system would comply with G83.

The FiT band is determined by the Total Installed Capacity (TIC) which has to be declared by the owner/installer. Normally this is based on panel sizes but the definition is open to interpretation. Once people at OFGEM start talking to me again (they seem to be it a bit of a huff with me at present) nailing the definitions of TIC and DNC down to sensible language is one of the next tasks.

Definitions of TIC and DNC are in this:

http://www.decc.gov.uk/assets/decc/...0100331172153_e_@@_fitlicencemodification.pdf

“Total Installed Capacity” means the maximum capacity at which an Eligible Installation could be operated for a sustained period
without causing damage to it (assuming the Eligible Lowcarbon Energy Source was available to it without interruption), a declaration of which is submitted as part of the processes of ROO-FIT Accreditation and MCS certified Registration;

“Declared Net Capacity” means the maximum capacity at which the installation can be operated for a sustained period without causing damage to it (assuming the source of power used by it to generate electricity was available to it without interruption) less the
amount of electricity that is consumed by the plant;

It's quite crazy that TIC is used for some purposes and DNC for others. And why does one say "could" and the other say "can"?

 

[TedM]

The system would comply with G83.

The FiT band is determined by the Total Installed Capacity (TIC) which has to be declared by the owner/installer. Normally this is based on panel sizes but the definition is open to interpretation. Once people at OFGEM start talking to me again (they seem to be it a bit of a huff with me at present) nailing the definitions of TIC and DNC down to sensible language is one of the next tasks.

Definitions of TIC and DNC are in this:

http://www.decc.gov.uk/assets/decc/what%20we%20do/uk%20energy%20supply/energy%20mix/renewable%20energy/policy/fits/1_20100331172153_e_@@_fitlicencemodification.pdf

“Total Installed Capacity” means the maximum capacity at which an Eligible Installation could be operated for a sustained period without causing damage to it (assuming the Eligible Lowcarbon Energy Source was available to it without interruption), a declaration of which is submitted as part of the processes of ROO-FIT Accreditation and MCS certified Registration;

“Declared Net Capacity” means the maximum capacity at which the installation can be operated for a sustained period without causing damage to it (assuming the source of power used by it to generate electricity was available to it without interruption) less the amount of electricity that is consumed by the plant;

It's quite crazy that TIC is used for some purposes and DNC for others. And why does one say "could" and the other say "can"?

 

[Worcester]

Thanks TedM,

With you having nailed the G83, looks like I've asked a question that's another can of worms! - Perhaps this next bit (FIT sizing / Declared Net / TIC) should be in a different thread?

From what you've said above, I would interpret that it is being limited by the inverter in this instance, especially as the definition refers to 'the installation' not the panels. - It would be good to get it nailed down!

If so this is good news for non south facing and more northerly located installations.

Gordon

 

[solarfred]

There is some confusion between G83 and FIT. The G83 limit of 16 amps is monitored by the DNO as they are responsible for the supply system and need to know who is putting what/how much, back in to the grid in any locality. The 16 amp limit is unambiguous because if you have a G83 approved inverter the output is capped at that limit otherwise it does not conform to the type testing certificate that should be included in the MCS installation paperwork. So the output of the inverter will be 16 amps /3680 watts @230 volts or 3840 watts at 240 volts. My DNO uses 240 volts and I have that in writing. They also are happy that the G83 inverter solves their problem as it is capped at 16 amps. What is installed on the roof is of no interest to them at all and is not their responsibility. The G83 installation certification clearly states it is DNC which is required as the output number.The installation diagram should identify the TIC (under OFGEM rules).
The 4 kw limit is dictated by the FIT system as OFGEM have decided that, that is the point at which the tariff payments drop from the highest rate to a lower rate. An installer wants to give the customer the maximum available output from the system and will normally try to install 4 kw on the roof to get as near to the 16 amp output for as long as possible. Using the limit of 3680w as the limit to install on the roof is completely wrong. The two limits are set for different reasons - one for safety and one for dictating how much gets paid out to customers under the FIT scheme. The confusion seems to be that you need more than 4 kw on the roof to get the maximum 16 amps from the inverter unless conditions are STC which is unlikely to happen in the UK very often. So to argue that it is the limit of the output of the inverter which dictates the total installed capacity is wrong as the limit of the FIT bands is not decided on the same set of rules (afaik) and has no bearing on the system output but only the FIT bands set by OFGEM. The inverter dictates the output of the system and this is recorded on the installation report but the TIC is dictated by the collectors on the roof and this is declared by the installer when commissioning the system. Two different things entirely. The MCS just oversee the whole process. They insist that the TIC or what is on the roof is "what is on the label" ie. add up all the wp of the installed modules and that is the TIC as they have been told by OFGEM that is the rule.
I think the TIC should be nearer 4400kw as that would be more likely to get maximum output from the inverter when conditions are right and this is the problem.Will OFGEM change the rules? I have serious doubts. It is not a safety issue just an issue about how much cash gets paid out and right now I don't think more cash is on the agenda.
If the FIT application form used TIC instead of DNC then it would be a lot less confusing and make slightly more sense but as Ted points out to use a mixture of terms across two different rules just makes a complete nonsense of the scheme. The rules were likely not written by individuals that have ever installed a PV system so would they not understand the implications of interpretation of the terms used.

 

[TedM]

The main problem, I think, is that a lot of these terms pre-existed outside of FiTs in other electrical regulations (terms used in regs that apply to large power stations which might self-consume large quantities of power, for example) - and they have just been reused in FiTs, I guess, as a shortcut. Hence why the wording seems so obscure and the application of the terms so tricky.

The main reason for the 4kW cut-off in rates is because this was seen as a threshold as far as the costs of installing a system was concerned; i.e. bigger systems have lower costs due to scale - as the whole intention of DECC was to keep a relatively flat rate of return across all system sizes. Whether 4kW was the exact right figure to use is impossible to say but, given the average size of domestic roofs, is pretty good.