Eskom ... is there ANY chance? In CPT there is

The shedule makes a lot of sense when there is sustained load-shedding over a longer period. It’s completely fair over 30 days.

But normal LS (now now excepted) is usually only a week, so some poor idiots get all the terrible slots in a row and others get “good” ones, or skip the bad slots due to stage changes.

In my opinion the schedule should actually be randomized, or at least not make patterns on the paper. That way it’ll be more fair when LS is only a week long.

But right now it’s pretty fair, since we’re shedding like a Husky in summer.

Can you imagine… My neighborhood whatsapp group would be swamped :wink:

You could shuffle it once, which would not strictly be random, but it would avoid the patterns. Basically take the 4 days where the same areas gets the same slot repeatedly, and spread them 6-8 days apart. Then reverse that to the other lower stages.

But even as I write that, I suspect that is going to be deeply unfair at lower stages. Instead of everyone getting their turn in a linear fashion, stage 1 and 2 will be completely haphazard, and a 2-3 day stint will disproportionately impact only some.

So if you want to keep the backwards compatibility with the lower stages (and you do!), you are stuck with what we’ve got now.

One thing that would solve the problem with recurring slots, would be to have a prime number of areas and/or slots. 12 and 16 are way too “round”.

Edit: And 12 and 16 have common divisors that are pretty much the cause of the recurring patterns. 17 Areas would make it shift nicely.

Edit 2: And stage 2, 4 and 6 being the most popular load-shedding levels also doesn’t help.

1 Like

Nah, can’t get more chaotic than it is now. I just keep posting a screenshot of COCT LS page and the URL, every time. I’ve stopped explaining, just post the URL again. And again, and again.

It will be less unfair than having LS for one week where one area has the morning traffic for the whole week, another the dinner-time slot for the whole week, etc. Also, short unfairness (2d) is easier to handle than a whole week.

But anyways, I’m sure there are other considerations as well.

I’m reminded of a second-year computer science class where the lecturer said that if you’re going to make a simple hash function for a hash table of sorts, using a modulo strategy with a prime number is a simple way to spread things more evenly. In this case, if we had 17 areas, it would totally solve the problem. But I suppose that’s a physical constraint. If we naively assume Cape Town works with 16GW, split into 16 x 1000MW blocks, then splitting it into 17 causes the blocks to be smaller, again complicating other math.

And as you said, during sustained periods like now, it ends up being perfectly fair. And during short stints, it is over before too many people start to complain. I mean… more than they usually do.

My go-to answer the last few weeks when people complain bitterly about LS:

It is going on 14 years that we’ve had bouts of LS, right? Yes yes yes yes the answer.
How’s the new car going/how was the holiday?
The complaining stops dead as the talk moves over to the new car/holiday.
… missing the point entirely, again.

Sometimes I ask them … Did you vote in every election since 1994?
Yes, the conversation continues.
No, the conversation is over. Go and B&M somewhere else, we’re busy here.

Some nice updates soon from ESP to make it clearer on what the stages/levels will be.

Saw ESP and wondered if ESP (Eskom se Push) is now going ESP (Extrasensory perception) to add “Upcoming” per area. :rofl:

Yeaaaah… I don’t know if you follow the CIAG people much, but their leading guy (Craig?) made a good point the other day. Since around 2006, the Western Cape has consistently NOT voted for the government we got. That old argument, that you have to work within the system to vote people out, isn’t working so well.

I’m not sure that the non-voters are really as guilty as we like to make them out to be. Sure, you’re shirking your civic duty, that much is true, but I am not sure you are disqualified from complaining. Too simplistic I think.

If you ask enough people, purely for your own interest, over a period of time, as I did, do, re. the 14 years of experience with not even a UPS and then sometimes add on the voting question, you may also see an interesting pattern emerging. :slight_smile:

Cannot report on my findings as it is an unscientific “study”, but it has in a sense been “peer-reviewed” kinda. :smile:

1 Like

I agree. You’re not disqualified. Abstinence is a form of complaint. They can complain, perhaps one just doesn’t take them too seriously for their lack of thought through on the matter. For many there’s only the ANC, so you either vote for them or not, and therein lies the rub

1 Like

Now if ESP adds Hz tracking to their app, then all and sundry can predict when loadshedding will start again…

1 Like

You can see the Hz of the grid on the live data available from your inverter (at least I can). The frequency is the same for everybody, it is what all the generators are synchronised on. If you see the frequency dropping fast… well then the proverbial paw paw has hit the fan. Loadshedding is there precisely to prevent the Hz from dropping in the first place.

2 Likes

Hz needs to be controlled between 49.85 Hz and 50.15 Hz, relatively tight control.

Groetnis

I monitored line frequency very accurately over a long period with custom hardware and could not detect any loadshedding precursors…

1 Like

I assume that if you can see it, then the grid operator already messed up. Load-shedding is supposed to prevent the Hz dip, not be a response to it. Their monitoring and alerting should be checking for earlier signs from the stations themselves I think.

Although, sometimes everyone has a very bad day.

1 Like

I have once (only once) seen a very pronounced dip below 50Hz just minutes before an Eskom announcement, but I concur with others here: When properly done, you won’t see the dip, because they are actively avoiding it.

Interesting indeed. You can see where I had loadshedding. Interesting other patterns while connected to the grid.

No way randomised could work in an industrial environment. We are a 24/7 blowmoulder that process HDPE thermally. Our production lines need to be at stable temperature before we can restart, so we lose a lot more than just the loadshedding hours. With the old 4-hour blocks we had half as many restarts, which was by far the preferred option compared to the new 2-hour stints. You can get a perpetual schedule for your area that repeats every month from Eskom somewhere - I hacked it to create our own schedule to accommodate the new invention of changing stages out of sync with loadshedding blocks (like at 5am resulting in 1-hour stoppages). Since mid-June we have lost 75% of our production hence 650kVA generator ordered - at least now we can use our 500kVA solar power plant during loadshedding as well. Spoke to our man at Nedbank today - he says manufacturing companies are requesting loans for 1) generators 2) solar 3) fuel-efficient trucks. Signs we are truly borked …

1 Like

I used the wrong word for what I meant: shuffled so there aren’t any patterns, then committed the same way it is now per day-of-month :slight_smile: