Two questions about solar panels.

[bernardgreen]

I suppose so, but it would perhaps be a little odd if compliance with a Standard or regulation were achieved by a secret process/mechanism, wouldn't it?

Most regulatory authorities only required only that the equipment functioned within the requirements of the authority, the means to achieve the functions were not examined. Primary concern is safety at the time of purchase.

Where the authority was also the customer with their own in house testing service then most often means were examined as well. The main interest was that components would have a minimum life time to ensure the equipment was reliable for its planned life time.

 

[ban-all-sheds]

I suppose so, but it would perhaps be a little odd if compliance with a Standard or regulation were achieved by a secret process/mechanism, wouldn't it?

Seems to be the way a lot of vendors of cryptographic products work....

 

[bernardgreen]

Seems to be the way a lot of vendors of cryptographic products work....

The people in Poland who invented the basic enigma machines for encoding commercially sensitive messages did explain how the machines worked to show potential users just how difficult it would be for any one to decode messages encoded by the machine.

The Germans decided it was safe to use thinking that even the Polish who developed it could never find the encryption key which was changed daily.

 

[JohnW2]

I suppose so, but it would perhaps be a little odd if compliance with a Standard or regulation were achieved by a secret process/mechanism, wouldn't it?

Seems to be the way a lot of vendors of cryptographic products work....

There are obviously some things which 'need' to remain secret (although, as Bernard has said, in the case of cryptography, many of the concepts of the process will often not be secret). However, that's not really the same as the sort of thing we are talking about here. If there are Standards and regulations for these inverters in order to avoid DNO personnel being killed, I think it fairly reasonable for it to be known, at least in principle, how compliance is being achieved.

In any event, I think this is probably all moot. Despite what westie suggested, I strongly suspect that any protection of IP associated with the design of these inverters will be achieved by patent protection, not 'trade secrets'. Not only is 'secrecy' a big gamble at the best of times but, in the context we're talking about, the inverters are freely available to be 'reverse engineered'.

Kind Regards, John

 

[JohnW2]

I guess they would never rely on something like that for critical shutdown signals.

I agree, but I remain intrigued to discover how they achieve assured shutdown under all circumstances ...

.. it is presumably far from impossible that there will be a segment of the network where a high proportion of consumers have solar systems and that, during periods of high sunlight and fairly low demand, that whole segment will be a net exporter of electricity to the grid. In that situation, it is far from obvious how any of the inverters could become aware of loss of the DNO supply to that segment. The solar installations will be more than able to serve the load, so there is no reason why their voltage should drop (or frequency change, at least in the short-term), and, on the face of it, all of them would be 'seeing' a situation which was no different from it would be would be were the DNO supply still there.

I suspect that there is a fairly simply answer that I'm missing.

Kind Regards, John

 

[AdrianUK]

Hmm backfeeding onto the DNO system during a power failure, as stated the inverters sense mains voltage and do shut down, ..

Indeed, but have you any idea how they do it? In other words, how do they sense loss of grid voltage when there is an inverter next door doing a good impersonation of the grid? Does it, as I've suggested, rely on the fact that the inverters alone will not be able to maintain a reasonable voltage in the presence of the likely loads, or what?

Kind Regards, John

There of several methods of detecting that you have lost your connection to the grid. Its not done simply by looking for a voltage - that would never work.

Firstly, a little theory... we consider the grid to be an 'infinite bus' ... its so large that nothing a small inverter or generator can do will change it & it will absorb all the power that the inverter can throw at it.

When a device is synchronised onto the bus it become locked to it .... any attempt to increase or decrease the system frequency will be opposed by large forces coming back from the grid. Let me turn to the case of a alternator being driven by an engine to illustrate. When the alternator is sync'd onto the grid, its rotational speed becomes locked by the frequency of the grid.... for a 50Hz system & a 4 pole alternator, that speed will be 1500 rpm. If you open the throttle of the engine & try to speed up the system, the engine rpm will not (cannot) change, but the torque transmitted across the mechanical coupling does increase. This torque is trying to rip apart the magnetic coupling inside the alternator - between the field on the rotor & the field created by the grid in the stator. This torque becomes converted into electrical power inside the alternator & flows out to the grid.

So what happens if the connection to the grid is lost? The system is no longer in equilibrium, the alternator is trying to produce electrical energy but there is nowhere for it to go so it must go into kinetic energy.... the machine set speeds up. The protection equipment will detect this rate-of-change-of speed (ROCOF) at the instant of separation & will act to shut down the generator.

Similar methods are used by grid tied inverters to look for the loss of equilibrium that occurs at the instant of separation. Another method is to look at the voltage vectors... at the instant of separation there will be a shift in the voltage vector as the system attempts to reach a new equilibrium state. This can be detected & used to disconnect.

 

[JohnW2]

Their of several methods of detecting that you have lost your connection to the grid. Its not done simply by looking for a voltage - that would never work.

Indeed - that's obviously been the point I've been making.

Firstly, a little theory... we consider the grid to be an 'infinite bus' ... its so large that nothing a small inverter or generator can do will change it & it will absorb all the power that the inverter can throw at it.

Agreed.

So what happens if the connection to the grid is lost? The system is no longer in equilibrium, the alternator is trying to produce electrical energy but there is nowhere for it to go so it must go into kinetic energy.... the machine set speeds up. The protection equipment will detect this rate-of-change-of speed (ROCOF) at the instant of separation & will act to shut down the generator. ... Similar methods are used by grid tied inverters to look for the loss of equilibrium that occurs at the instant of separation.

Are you saying that these grid-tied inverters, unlike inverters supplying equipment, do not have any built in regulation? If I took one of the latter and reduced (or even removed) its load, if it were a decent one, neither the voltage nor frequency of its output would increase - it would merely draw less power from the source (hence there would be no energy which 'had nowhere to go'). I suppose that, even with such regulation, signals from the regulation circuitry could be used to indicate that the 'load' ('sink capacity' of what it was seeing of the network) had suddenly changed.

However, given the need for an ultra-certain system, I would see one worrying catch with that sort of sensing. If, by pure chance, at the moment of 'separation', the total load still presented by the 'separated' part of the network were very close to the output capacity of the attached inverters, they would presumably not see any change in anything, would they? A very improbable scenario, I admit, but not impossible.

Another method is to look at the voltage vectors... at the instant of separation there will be a shift in the voltage vector as the system attempts to reach a new equilibrium state. This can be detected & used to disconnect.

Again, would that necessarily be the case in the scenario I have mentioned in the preceding paragraph?

Kind Regards, John

 

[ban-all-sheds]

There are obviously some things which 'need' to remain secret (although, as Bernard has said, in the case of cryptography, many of the concepts of the process will often not be secret).

Actually with cryptography it is now absolutely clear that the entire development process need to be open - open source code, open standards, openly developed standards - not one single aspect of how the algorithms work or how the code is developed and tested, of how the design is done etc, can be kept a secret.

Basically the US Govt have compromised everybody - none of us now have any guarantee of privacy, none of us now have any guarantee of security in our commercial transactions and communications, and that applies all the way up from an individual using his credit card to transactions and communications which can make or break the company he works for, his pension fund, even the country he lives in.

The NSA and other related agencies could not have done what they did if the entire process for creating cryptographic products was open.

 

[Lectrician]

So, to simply Adrian's great post........the inverts generally lock to the frequency, and run with it. Mains supply goes, the frequency being followed goes, inverter shut down.

This is basically what I have been told.

 

[ericmark]

I had wondered when I wanted inverter power in a boat if I could increase the power of the existing inverter by adding a grid tie inverter but was assured it would not work.

I would think even with a street of 100 grid tie inverters with the loss of the grid either the frequency will change or the voltage will move out of spec so the inverters will close down.

However I thought the inverters would auto restart so unless all failed before one tried to restart it may have still continued to work.

However today the guy who was going to install the solar panels arrived and told me there was no auto restart and after it drops out the house holder has to go through a re-start procedure. So although time for all to drop out may increase as more and more inverters are added it will still go into lock down mode it just may take a little longer.

 

[DetlefSchmitz]

... after it drops out the house holder has to go through a re-start procedure....

I never have had to.

 

[westie101]

and told me there was no auto restart and after it drops out the house holder has to go through a re-start procedure.

Not that I'm aware of, there may be a time delay but I believe they do auto-restart.

 

[Lectrician]

Isn't it in the spec, 2 seconds to disconnect and shutdown on power fail, 3 mins before reconnect after power returns. They all auto start.

Even when they shutdown for other reasons, such as over voltage etc, they will restart the self.

 

[dormermike]

Encrpytion methods have been standardised and open to scrutiny for years, well before any of this NSA malarky. Worst thing a software dev can do is roll their own 'encryption'.

 

[JohnW2]

So, to simply Adrian's great post........the inverts generally lock to the frequency, and run with it. Mains supply goes, the frequency being followed goes, inverter shut down. This is basically what I have been told.

Not as I understand it. I read him as essentially saying that, by analogy with a poorly-regulated motor/alternator set up, loss of 'anywhere for the energy to go' would result in the motor speeding up (or trying to speed up) and that change in frequency would be detected and used to result in a shutdown. However, as I explained, I could not really see how that could be totally reliable.

In any event, if what you say above were the case, I still don't really understand how it would work (in the presence of other inverters). The inverter will lock to the frequency (and phase) of 'what it sees'. Normally, all the inverters and the grid supply are therefore locked together. At the moment of grid failure, each inverter will still see the output of the other inverters, and will simply continue to lock to them, won't it?

Kind Regards, John