Loadshedding early warning

[JTech]

I think I have it for an early warning of imminient load shedding

As you are aware I'm doing research on the ripple control systems used in JHB I am working on decoding the data sent and what I have found is that today, some very peculiar data being sent, where basically everything is being turned off across the board. I have of course seen this before and when I did see it last time, I happened to be working on the project and the power went out very shortly thereafter.

I saw a similar transmission this morning so, I am now idling here and waiting to see what the exact sequence is before the power is cut, I believe it will be cut at 10:00 which is exactly 2 hours away so let's see if it works out.

The benefits of this is huge, we'd be able to switch a generator in without power dipping or interrupting the supply to sensitive stuff like computers and CNC machinery. No UPS needed in other words.

[Dave A]

We had the pleasure of being load-shedded yesterday, so it's definitely happening.

It's interesting that you can monitor the signals being sent out to the ripple relays. Are you watching for patterns in when the relays are activated?

[JTech]

Yes it is happening. I went to the office after seeing that pattern this morning and at exactly 10h00, as expected, the power was cut. Without looking at the schedules I could see it coming, very very handy that.

In general I can decode (on paper presently- am debugging my code still) the actual transmission. What I do see is that with the loadshedding, seemingly unused channels suddenly become alive...

The transmission code was a closely guarded secret until I broke the code and confirmed my findings with Orionpower in NZ who are generally happy to share their ripple control codes.

The manufacturer, and it's agent in SA has a rather crap attitiude and were of no help either, they actually gave me the raised middle finger and threatened me but as far as I can see, the code is not patented and trust me, I did exhaustive searches on IEEE and various patent websites.

I currently extract the ripple into a bitstream using a digital filter algorithm I developed myself. I then capture the bitstream with a digital storage oscilloscope, and print it out and then count the bits and decode them on the page.

[duncan drennan]

YThe manufacturer, and it's agent in SA has a rather crap attitiude and were of no help either, they actually gave me the raised middle finger and threatened me but as far as I can see, the code is not patented and trust me, I did exhaustive searches on IEEE and various patent websites.

Is that Farad that you are referring to?

[Snoopy_inc]

I would love if you could post that code onto my website if you get a chance

My members thrive on said things

Thanks Jtech

Site is in my signature

[Dave A]

I can understand the concern for protecting the codes. It wouldn't take that much to inject it into the supply.

It would only get as far as the transformer, but that could be enough to wreak some havoc, I'm sure.

[SilverNodashi]

We had the pleasure of being load-shedded yesterday, so it's definitely happening.

We also had load shedding yesterday, 10am exactly.

It's interesting that you can monitor the signals being sent out to the ripple relays. Are you watching for patterns in when the relays are activated?

What can one do with this kind of info though?

[Dave A]

What can one do with this kind of info though?

Load shedding is a problem in its own right. But the problem is being aggravated by it being somewhat unpredictable.

We're being told when we are going to be load-shedded if the need arises. So that might seem OK.

But a business prepares to be load-shedded by doing some rescheduling because it seems load-shedding is likely - and then it doesn't get load-shedded after all - that can also cost money.

For example, Ian could plan a print run to make sure the power doesn't go down in the middle of the run, and accepts some inefficiencies to make sure that when the power goes off at (say) 10.00 am, the machines are idle anyway. Now if the power stays on, of course he hasn't lost those 2-4 hours, but he's still lost the time making sure everything is clear at a time when his machines would normally be running full steam ahead.

It might be less of a disaster than grinding to a halt in the middle of a run, but it's still a loss situation.

[SilverNodashi]

that makes sense

[JTech]

Hello all

Eskom has found my website and consequently I have received correspondence regarding my project, from their staff who deal directly with the ripple control system.

One of the senior managers there was kind enough to e-mail me two very valuable PDF documents detailing the ripple control system.

I was not far off the mark with my research, I was in fact very close to cracking the protocol.

Rest assured that the information now at my disposal will accelerate this project thousand-fold. It also dispels the myth that the protocol is "proprietary" and is actually in the public interest.

That aside, if all goes well, the systems will be marketed towards the end of September and include generator control and if Eskom agrees to my suggestion, some sort of national load status indication on a spare channel.

Okiedokie I am off, got a lot of work to do. Will report back in =+/- 3 weeks

Cheers for now
JTech