Jump to content

Reverse Powerflow Protection


Recommended Posts

Decided to try using a RHOMBERG SP510 to monitor RPF, and hopefully prevent the LANDIS+GYR prepaid meter from tripping. A 100A ct will be fitted to the live grid supply cable. The SP510, will control a 40A 4-pole N/O mechanical contactor, where my incoming grid load will be bridged across 2 contacts each. I know electrical engineering design, requires double contacts, decreasing the failure factor.

As I am not in the COCT area, and don't have any local requirements to adhere too, I wanted to see how effective this will be. I do not intend exporting to the grid, so just wanted to add an extra level of safety in respect of ISLANDING, and the chance that these requirements, may come into effect some time in the future.

A PDF spec sheet of the SP510, and pictures are attached. Still need to figure out how exactly to get it installed in the most practical way tomorrow.

 

IMG_5749b.jpg

IMG_5750b.jpg

Rhomberg SP510.PDF

Link to post
Share on other sites
1 hour ago, Blinkwater9 said:

and hopefully prevent the LANDIS+GYR prepaid meter from tripping.

I am not sure what Landis+Gyr you have that will trip on reverse current and what the response time of the prepaid is to reverse current, but keep the flowing in mind. 

Conlog and Ecolec pre-paid Meters will disconnect from the grid within 15ms if the reverse flow current is above a certain level. I wanted to use the SP510 to try and prevent this from happening. As can be seen from the spec sheet,  the response delay is adjustable from 1 to 10 seconds. Before the Rhomberg can react, the two mentioned pre-paids would have tripped already.  

It would still be interesting to to see the result. Please keep us updated?

Rom.thumb.JPG.afd2bb2bbc6dc5b80c3922a7ac3aaf83.JPG

 

Link to post
Share on other sites
9 hours ago, Jaco de Jongh said:

I am not sure what Landis+Gyr you have that will trip on reverse current and what the response time of the prepaid is to reverse current, but keep the flowing in mind. 

Conlog and Ecolec pre-paid Meters will disconnect from the grid within 15ms if the reverse flow current is above a certain level. I wanted to use the SP510 to try and prevent this from happening. As can be seen from the spec sheet,  the response delay is adjustable from 1 to 10 seconds. Before the Rhomberg can react, the two mentioned pre-paids would have tripped already.  

It would still be interesting to to see the result. Please keep us updated?

Rom.thumb.JPG.afd2bb2bbc6dc5b80c3922a7ac3aaf83.JPG

 

I am fully aware of the response time, and have looked at other options too. My main goal is to disconnect from the grid, in the event of a power failure or load shedding. I tried contacting the manufacturer of the prepaid meter, but just got a generic response. The SP510 appears to be the only device available which is designed specifically for this function? I havent been able to find anything else, which clearly states the function of REVERSE FLOW MONITORING?

Maybe someone can advise on a faster alternative.

An electrical engineer I spoke to, seemed unfamiliar the the terminology, and suggested 4-quadrant metering, but the specification documents of one of these meters I looked at, does not even mention the function, and though I am sure the meter can monitor or measure reverse power flow, I dont think it can react to it, at least not any better than the SP510?

Comments and opinions welcome?

 

Link to post
Share on other sites
10 hours ago, Jaco de Jongh said:

I am not sure what Landis+Gyr you have that will trip on reverse current and what the response time of the prepaid is to reverse current, but keep the flowing in mind. 

Conlog and Ecolec pre-paid Meters will disconnect from the grid within 15ms if the reverse flow current is above a certain level. I wanted to use the SP510 to try and prevent this from happening. As can be seen from the spec sheet,  the response delay is adjustable from 1 to 10 seconds. Before the Rhomberg can react, the two mentioned pre-paids would have tripped already.  

It would still be interesting to to see the result. Please keep us updated?

Rom.thumb.JPG.afd2bb2bbc6dc5b80c3922a7ac3aaf83.JPG

 

Even the ZIEHL is rated at 0,05 seconds, which is 50ms by my calculations, and my maths is terrible, lol. I think the only wat, to prevent the prepaid meter from tripping, is to have a meter installed that wont trip, and this is a 💩pile I am not prepared to step in?

Link to post
Share on other sites
7 minutes ago, Blinkwater9 said:

Even the ZIEHL is rated at 0,05 seconds, which is 50ms by my calculations, and my maths is terrible, lol. I think the only wat, to prevent the prepaid meter from tripping,

I am sorry, but this part I dont understand, the Ziehl is designed to measure Over and Under Voltage, Over and Under Frequency and Vector Shift n three phase connections. That is why it would be able to sense disconnection from the grid  in the event of a power failure and then insure that the contractors its connected to will de-energize. Its not to detect and disconnect from the grid in the event of reverse flow. 

I can sort of understand what you are trying to achieve, but dont think it is advisable to  disconnect from the grid every time you sense reverse flow. I would not switch the AC to my inverter off and on multiple times a day.  

Link to post
Share on other sites
22 minutes ago, Jaco de Jongh said:

I am sorry, but this part I dont understand, the Ziehl is designed to measure Over and Under Voltage, Over and Under Frequency and Vector Shift n three phase connections. That is why it would be able to sense disconnection from the grid  in the event of a power failure and then insure that the contractors its connected to will de-energize. Its not to detect and disconnect from the grid in the event of reverse flow. 

I can sort of understand what you are trying to achieve, but dont think it is advisable to  disconnect from the grid every time you sense reverse flow. I would not switch the AC to my inverter off and on multiple times a day.  

Hi Jaco,

My main goal is to try to prevent the prepaid meter from tripping, and resulting in issues with the local ESKOM croud if such an event occurred? I agree that disconnecting the AC GRID supply is not a good idea, I hope to not have this happen? Hopefully, and I cant confirm how this unit will react, I am able to prevent reverse power flow in a worst case scenario only?

I am also convinced, that this issue will arise again in the future, as local legislation catches up? 

Has anyone actual had a ZIEHL RELAY prevent reverse power flow where the reaction has been witnessed as such? Or seen this unit in a test application, where the function can be tested?

This appears to be another one of those topics, where in-depth technical information seems hard to find? Although there are a few units on offer being sold as REVERSE POWER FLOW PROTECTION or ANTI-ISLANDING devices, the technical information seems a little unclear, to me at least? Maybe my lack of experience in this field is my downfall here?

Link to post
Share on other sites

I am not familiar with this device per se, but I am expert on protection relays in general.

Short of putting these meters through their paces, (which wouldn't be too difficult), I believe I can bring some common sense into this discussion.

I think it must be understood more in depth what tripping in 15ms means and what it does not mean.

Well firstly 15ms is pretty fast to trip to start with, but possible. Practically, this would introduce more expensive parts into a meter, but maybe they are there. Lets accept this spec at face value. bear in mind though there is a cost associated with sophistication, and sophistication can be a requirement in places, but this is just a meter.

Secondly this time  will only describe the trip time once a trip decision is made.

Why do I say this, well, 15ms is too fast to measure, make a trip decision and carry out and complete the action. Way too fast for reverse power protection. This level of sophistication in a mass produced meter ( read cost), why? I don't see it being practical.

Reverse power protection will measure the phase angle relationship between voltage and current over X number of cycles, each cycle being 20ms. (50hz)

The device will probably integrate or count up to Y consecutive cycles of reverse power flow and then only issue a trip. This will take multiples of 20ms.

There are other system considerations that such a meter must remain stable for, such as motors that briefly will become generators during power system dips and the switching of large inductive loads that would require a meter ( which isn't a protection device) to have some practical ride through capability. Otherwise these meters would be too trigger happy to be usable.

Now that Rhomberg relay may just be too coarse and slow, but maybe not. Try it if you have it, use the fastest time and highest current sensitivity.

 

Link to post
Share on other sites

The conlog BEC23 will allow 40W for 15 seconds, so for lower levels it will wait longer, at 600W it will trip within a second. Most meters have to provide a margin to prevent nuisance tripping, so they are not going to trip within milliseconds on just a few watts (though they may once you're up in the kilowatt range).

The BEC44 is twice as sensitive as the BEC23. I have no idea what other meters use. To avoid tripping one has to make the feed-in event low enough or short enough. You don't necessarily have to completely prevent it.

The second item I wanted to address is not to confuse anti-islanding with reverse blocking. Reverse blocking is to prevent power from flowing backwards into the grid while the grid is live. Anti-islanding is about preventing flow into the grid while the grid is NOT live. It's actually completely different areas of concern. The relay in question cannot prevent islanding, and as I recall, it's actually a power factor relay and you're relying on the fact that power factor goes negative when you feed back.

The third thing, is to remember that when a hybrid is balancing power at zero, the power factor gets absolutely atrocious (where measured at least), I've seen my Carlo Gavazzi log PF of 0.04 when it's hovering around the zero point. In other words, I foresee this actually causing nuisance tripping.

Finally, these things usually have a sensitivity that's proportional to the CT you use. With a 100A CT it might not be able to measure that well low down.

But let me suggest an alternative, which is what I had in mind many months ago when I also looked at this, and that is to somehow (not sure how yet) use the relay to turn ON a load when the power factor goes out of bounds. The idea is simply to turn something on to absorb the feed-in event.

Link to post
Share on other sites
1 hour ago, plonkster said:

But let me suggest an alternative, which is what I had in mind many months ago when I also looked at this, and that is to somehow (not sure how yet) use the relay to turn ON a load when the power factor goes out of bounds. The idea is simply to turn something on to absorb the feed-in event.

Yes, I was thinking the same thing, I actually edited it out of my post, but seeing as you brought it up.

If the Rhomberg doesn't suit the conventional reverse power connections:

I'd try  reversing either the current or the voltage connections to the relay, so that it became a forward power relay, then I'd use the NC contact ( which would be open as the relay would normally be drawing sufficient power the grid and in the tripped condition) to bring in a load when the forward power got sufficiently low. ( as the relay reset).

This should theoretically give the relay a head-start on the meter for a a non-step change increase in embedded generation.

Link to post
Share on other sites
1 hour ago, phil.g00 said:

reversing either the current or the voltage connections to the relay

It uses a CT. Literally just flip the CT over to make the current 180 degrees out of phase with the voltage 🙂

Link to post
Share on other sites
On 2019/06/29 at 7:41 AM, Blinkwater9 said:

I am also convinced, that this issue will arise again in the future, as local legislation catches up?

i agree. I believe that we will, like most other countries, eventually end up with (reasonably priced) bi directional meters on all GTI's. In the mean time the municipalities and Eskom should immediately provide the tokens to prevent nuisance tripping of the one way meters. @Rautenkdid this issue come up in your discussions with the CoCT? What is Reyno's view on this? 

Link to post
Share on other sites
On 2019/06/28 at 6:16 PM, Blinkwater9 said:

Decided to try using a RHOMBERG SP510 to monitor RPF, and hopefully prevent the LANDIS+GYR prepaid meter from tripping. A 100A ct will be fitted to the live grid supply cable. The SP510, will control a 40A 4-pole N/O mechanical contactor, where my incoming grid load will be bridged across 2 contacts each. I know electrical engineering design, requires double contacts, decreasing the failure factor.

As I am not in the COCT area, and don't have any local requirements to adhere too, I wanted to see how effective this will be. I do not intend exporting to the grid, so just wanted to add an extra level of safety in respect of ISLANDING, and the chance that these requirements, may come into effect some time in the future.
 

Hi Blinkwater9

I am not sure which Landis & Gyr meter you have, but I have the "CIU3" 3-phase meter and the option for tripping on "brief reverse energy" is turned off by default. 

This is described in Meter Register 35 - see attached doc.

Landis & Gyr CIU3 Electricity Prepayment Meter.pdf

Edited by NigelL
Link to post
Share on other sites
On 2019/07/02 at 3:38 PM, phil.g00 said:

I am not familiar with this device per se, but I am expert on protection relays in general.

Short of putting these meters through their paces, (which wouldn't be too difficult), I believe I can bring some common sense into this discussion.

I think it must be understood more in depth what tripping in 15ms means and what it does not mean.

Well firstly 15ms is pretty fast to trip to start with, but possible. Practically, this would introduce more expensive parts into a meter, but maybe they are there. Lets accept this spec at face value. bear in mind though there is a cost associated with sophistication, and sophistication can be a requirement in places, but this is just a meter.

Secondly this time  will only describe the trip time once a trip decision is made.

Why do I say this, well, 15ms is too fast to measure, make a trip decision and carry out and complete the action. Way too fast for reverse power protection. This level of sophistication in a mass produced meter ( read cost), why? I don't see it being practical.

Reverse power protection will measure the phase angle relationship between voltage and current over X number of cycles, each cycle being 20ms. (50hz)

The device will probably integrate or count up to Y consecutive cycles of reverse power flow and then only issue a trip. This will take multiples of 20ms.

There are other system considerations that such a meter must remain stable for, such as motors that briefly will become generators during power system dips and the switching of large inductive loads that would require a meter ( which isn't a protection device) to have some practical ride through capability. Otherwise these meters would be too trigger happy to be usable.

Now that Rhomberg relay may just be too coarse and slow, but maybe not. Try it if you have it, use the fastest time and highest current sensitivity.

 

Awesome information, Phil. Thank you for taking the time to share this information with us. I do have the unit, and have already built the basic setup, and will be installing it next weekend.

If I have any new information, I will share it on this thread.

Link to post
Share on other sites
On 2019/07/03 at 10:27 AM, NigelL said:

Hi Blinkwater9

I am not sure which Landis & Gyr meter you have, but I have the "CIU3" 3-phase meter and the option for tripping on "brief reverse energy" is turned off by default. 

This is described in Meter Register 35 - see attached doc.

Landis & Gyr CIU3 Electricity Prepayment Meter.pdf 2.02 MB · 1 download

Hi Nigel,

My LANDIS+GYR is a SABRE ED. Interestingly, TOKENS are mentioned in this thread too? In this PDF, there is information listing tokens? Not quite ready to fiddle with the meter yet?

Maybe someone, who has had these tripping issues, can share the process of resting or restoring the meter? How the process works? Maybe being familiar with this process, can help us, should we find ourselves in this situation?

Thanks,

User Guide Sabre Mk5 Rev 0.0.pdf

Link to post
Share on other sites
On 2019/07/06 at 11:30 AM, Blinkwater9 said:

In this PDF, there is information listing tokens?

The token needed to enable/disable the reverse power feature is generally not provided to the public. You need to beg the supplier to give it to you. The "support" guy at your supplier has likely never heard of this problem, doesn't know what to do, won't be told how to do his job, and is more concerned with other matters.

🙂

 

Edited by plonkster
Link to post
Share on other sites
  • 2 weeks later...

Just an update. Installed the REVERSE POWERFLOW PROTECTION RELAY, Rhomberg SP510, and an 80Amp 230vac contactor. The original picture shows a 50amp 4-pole contactor, but

I had an issue getting the 10mm² cables into the terminals, so I increased the amp rating to the 80Amp version, and instead of doubling the the contacts, just used the 3-pole version.

I also purchased a CT closer to the rated 63amp supply. Tested the unit by inverting the CT, and it definitely trips out the contactor. I'm sure this will suit my need, for an anti-islanding device, should this need arise. The delays and sensitivity are adjustable, and I intend fiddling with these somewhere down the line.

I also increased the size of the enclosure, giving me more space for the heavier cabling. I also left the 50amp contactor on the rail, should it be required later.

 

IMG_5995a.jpg

Link to post
Share on other sites
1 hour ago, Blinkwater9 said:

for an anti-islanding device

Practically it would probably do the trick, but I'm pretty sure it has a huge NDZ (non-detection zone). I mean, assume the grid has failed and my inverter failed to disconnect, and I'm inadvertently feeding 50mA into my feeder line. That's enough to kill someone, but it's only 10W going in the wrong direction.

Still, impressive use of commodity hardware, well built, and like the bootlaces.

Link to post
Share on other sites
5 minutes ago, plonkster said:

Practically it would probably do the trick, but I'm pretty sure it has a huge NDZ (non-detection zone). I mean, assume the grid has failed and my inverter failed to disconnect, and I'm inadvertently feeding 50mA into my feeder line. That's enough to kill someone, but it's only 10W going in the wrong direction.

Still, impressive use of commodity hardware, well built, and like the bootlaces.

Thanks. The bootlaces do make a neat build. You should see my wire markings?LOL. Wrap-around thermal labels, on all the wires. My thinking here was, should I need to move something from Non-essential to Essential load, I would need to move the neutral and earth too, as each is seperated. This way, I don't need to search for wires.

 

To be quite honest, I'm not concerned over a momentary feed-in or spike. What I want, is should the grid go down for any reason, I want the disconnection to be made.

Also, I imagine, should any compliance be required later, I seriously doubt that anyone will be able to confirm that the this unit would not fulfill the requirement of an ANTI-ISLANDING device. 

And on the point of tripping the pre-paid meter. I am lead to believe that on my meter, the feature to shut down on reverse power detection, is disabled.

I think, if legislation requires a device to prevent REVERSE POWER FLOW, then at +-R2000, this unit should fulfill the requirement. The way I understand this issue, there are 2 clear terminologies that are used? REVERSE POWER FLOW DETECTION, and ANTI-ISLANDING. Reverse power flow detection/prevention, should react to prevent the electricity meter from seeing this as an attack, and Anti-Islanding is isolation from the grid, to protect maintenance workers from electrocution?

Lets see some opinions?

 

 

IMG_5476[1].JPG

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...