Regulation 551.7.2 - Generating Sets Operating in Parallel with Grid Fed Electrical Supplies (specifically solar PV related)

Question

I know there have been a few discussions based on this but none I can&rsquo;t find relating to a commercial three-phase roof mounted solar PV installation. 
 With regards to regulation 551.7.2, do you have to use the rated AC output current of inverter(s) for Ig(s), or can you take an engineered approach based on what can actually be generated at any one time? 
 
 InA &ge; In + Ig(s) where: 
 InA is the rated current of the assembly (consumer unit or distribution board for e.g.) 
 In is the rated current or current setting of the incoming circuit overcurrent protective device either incorporated within the low voltage switchgear and control gear assembly or upstream of it 
 Ig(s) is the rated output current of the generating set or sets. 
 
 If you had 2 inverters with a maximum rated AC output of 200A 3PH each as stated on manufacturer data sheet, you would say that Ig(s) is 400A 3PH total for an array with all PV modules facing in the same direction. 
 However, if half of your PV array was facing East and was connected to one inverter, and the other half connected to the second inverter was facing West, you would never generate the maximum rated output at any one time, as it would be reduced to 200A approx. 
 Therefore, could you use this lower figure of 200A for Ig(s) to satisfy this regulation or is it explicit that the combined maximum rating of 400A for all inverters should be used, even if it couldn&rsquo;t be reached? 
 My concern is that LV assemblies may need to be rated to a higher current that in reality, would never be present due to the managed design of the PV system. 
 Also, that existing installations whose LV assemblies were built before solar PV was widely used, could not benefit fully from an appropriately sized solar PV installation. 
 Thanks 
 P.S. 
 The above situation is based on the 2 inverters being cabled back to a dedicated 3PH MCCB sub LV panel board, which is then cabled back to a 3PH main LV switchboard at the intake. 
 I do realise one fix could be to set down the protective device upstream of the main LV switch board however it would take some convincing of the client to do so, even though I know their load profile would allow it! So please humour me and disregard that option.

Michael Peace · Accepted Answer

I hope this is useful. 
 wiring-matters-november-2024.pdf (theiet.org)

gkenyon · Answer

Electron666 said: With regards to regulation 551.7.2, do you have to use the rated AC output current of inverter(s) for Ig(s), or can you take an engineered approach based on what can actually be generated at any one time? 
 Well, I don't understand this question. Because if the inverter has to be rated to output the total load of all panels, why not the connecting circuit, and the distribution board that the inverter connects through? What i mean by that, is that if Ig(s) is the AC output of the inverter, then surely you address that by the rating of the inverter (if it really can only output maximum what the panels can provide at any one time ... why would you put an over-rated inverter in?)

mapj1 · Answer

That is not how I read it - to me it seems more like there are 2 gensets but mechanically interlocked by sharing a common source of sunshine, so that only one of them is ever at full power at once.

but that is also something that the wording is not quite intended for....
Mike

gkenyon · Answer

Electron666 said: I agree with this in situations where all PV modules face the same way, however in my case of East-West, 400A will make In+Ig(s) greater than InA, which of course is non-compliant. 
 So my question is, do you have to sum the AC output of all inverters together to determine Ig(s), or for an East-West situation can you apply engineered logic to arrive a lower, more accurate and no less safe figure? 
 If this engineered approach was deemed acceptable, would it need to be documented by say a derogation from this particular regulation and / or a risk assessment? Thanks 
 To answer this, I would have to "interpret" BS 7671, which I'm not permitted to do ... in fact, only a court of law can do that. You could seek a professional opinion (or indeed hold your own), but if anything were to go awry, it would still be open to scrutiny in court? As such, I think all anyone could offer in response to this question at this stage, is that if, and only if, you (and if applicable your PI insurer) are happy to rely on the modelling providing a particular maximum of I g(S) , it seems like a reasonable approach. It all, I guess, rests on the reliability and caveats involved in the modelling ... would it stand up to external scrutiny ? 
 I think that a better approach to improving on the validity of the assumption would be to include curtailment, so that if your modelled/calculated I g(S) , is exceeded, one of the inverters is curtailed accordingly? This would also align with one of the methods put forward in Section 9.3.4.2 of the IET Code of Practice for Electrical Energy Storage Systems, 3rd Edition, for meeting Regulation 551.7.2 of BS 7671?

AJJewsbury · Answer

Indeed but InA &ge; In + Ig(s) gets a lot easier to satisfy if you can take In to be zero (since no load will ever be drawn from the grid the rating of the grid side fuse is immaterial to the loading on the DB). 
 - Andy,

gkenyon · Answer

AJJewsbury said: The 2nd part is the same idea but for DBs. The OP does raise an interesting point though - if a DB serves generators alone, and no loads, common sense says that it only needs to be rated for I g(s) alone - but the wording doesn't seem to take that possibility into account. 
 This is where some of the other options put forward in the IEC CoP for Electrical Energy Storage Systems, 3rd Edition, come into play. We saw something similar in the Draft for Public Comment of the proposed Amendment 4 to BS 7671:2018 ... so, whilst at present BS 7671 only has the single formula, we know it is at least currently "under review".

Simon Barker · Answer

Have you worked out what the generation will be around midday on a sunny June day? The sun will be directly overhead, so illuminating both sets of panels. Depending on the mounting angle of the panels, you could get something close to the kWp rating from both sets of panels at he same time.

mapj1 · Answer

551.7.2 A generating set used as an additional source of supply in parallel with another source shall be installed: &ndash; on the supply side of all the overcurrent protective devices for the final circuits of the installation, or &ndash; on the load side of all the overcurrent protective devices for a final circuit of the installation, but in this case all the following additional requirements shall be fulfilled: (i) The conductors of the final circuit shall meet the following requirement: Iz &ge; In + Ig where: Iz is the current-carrying capacity of the final circuit conductors In is the rated current of the protective device of the final circuit Ig is the rated output current of the generating set (ii) A generating set shall not be connected to a final circuit by means of a plug and socket-outlet (iii) The line and neutral conductors of the final circuit and of the generating set shall not be connected to Earth (iv) Unless the device providing automatic disconnection of the final circuit in accordance with Regulation 411.3.2 disconnects the line and neutral conductors, it shall be verified that the combination of the disconnection time of the protective device for the final circuit and the time taken for the output voltage of the generating set to reduce to 50 V or less is not greater than the disconnection time required by Regulation 411.3.2 for the final circuit. 
 
 The reg is not really worded with your east and west panels type situation in mind. The option that absolutely always complies but is a waste of money and materials is the one where you assume your two banks of 200A are in fact capable of 400A of simultaneous generation, even though it cant happen. Another approach may be to common the two banks, into a single 200a fuse per phase, and then have that connected at a single point as a 200A max limited total generation. Direct connection may be shown by calculation to always be safe, but you may leave yourself open to some pedantic type late claiming it does not comply with the regs as written, even if it does comply with the regs in spirit. The other complication, at least for a large installation with lots of origins, is that this really only applies to installations where solar is connected to an existing "final" circuit (eg presumably one that has consuming devices also wired into it). It is not so clear if it applies to "dedicated" generating circuits without consuming devices - presumably as the overload situation cannot occur if there is not a source and load that connect direct to each other without passing through a fuse or breaker first.. 
 It does make you wonder a bit exactly how this reg works in practice when you see what happens when things like this get installed. plug-in 4.67kw-diy-solar-power-kit Or even this battery unit advertised by EDF 
 looks just 'plugged in ' to a UK socket from this video ( https://youtu.be/BHBp_XHQ0rE) Mike.

AJJewsbury · Answer

551.7.2 A generating set used as an additional source of supply in parallel 
 
 I suspect the OP is more concerned with the 2nd part of the reg (possibly new to AMD 3) ... "Where the generating set is connected to a low voltage switchgear and controlgear assembly "... requiring I nA &ge; In + I g(s) . 
 The first part of the reg only applies when the generator is connected to a final circuit that has other loads connected to it as well (hence the loads on that circuit could draw In from the mains plus Ig from the generator without troubling any over current devices) - the more usual situation is that the generator has its own circuit (which as it has no loads or socket outlets is by definition not a final circuit) so the the generator connection meets the 'on the supply side of all overcurrent protective devices for final circuits' bit, so the rest of the 1st part doesn't apply. 
 The 2nd part is the same idea but for DBs. The OP does raise an interesting point though - if a DB serves generators alone, and no loads, common sense says that it only needs to be rated for I g(s) alone - but the wording doesn't seem to take that possibility into account. 
 - Andy.

Regulation 551.7.2 - Generating Sets Operating in Parallel with Grid Fed Electrical Supplies (specifically solar PV related)

Join us to get the best from IET EngX.

Community Rules & Guidelines