Octopus Saving Session

badger09

Thanks to those who posted info about part units counting. I couldn’t see anything in the FAQs & don’t tweet. 

@wild666 I’m planning to switch off fridge freezers for the hour tonight just as a trial. Don’t think I’d do it for 3 hour sessions though. 

Petriix

Deleted_User said:

Petriix said:

One thing which remains unclear is the notion of an 'In Day Adjustment' which they haven't clearly defined. The most obvious interpretation is that they will adjust the average of the previous 10 days' usage for that hour based on the relative use on the actual day so, if your overall usage on the day is higher than average, your relative saving will be considered to be higher (and vice versa). However, it could be the exact opposite of that and it really isn't clear at all.

I wouldn't expected it to be your relative use that gets compared for the in day adjustment - rather the general system demand or some sort of regional number.

Sort of "you usually use 2kWh, but everyone is using 50% more than normal today so we'll compare your saving against a 3kWh target".  This could end up with a stable use being counted as a saving - i.e. your usage didn't go up when everyone else's did.

Certainly could do with some clarification.

It very specifically is 'you', it's just not clear how the adjustment is calculated and what effect it has.

These half hourly averages are then subject to an ‘In Day Adjustment’ which will make adjustments based on how much energy you’ve used on the day of the session.

[Deleted User]

Petriix said:

Deleted_User said:

Petriix said:

One thing which remains unclear is the notion of an 'In Day Adjustment' which they haven't clearly defined. The most obvious interpretation is that they will adjust the average of the previous 10 days' usage for that hour based on the relative use on the actual day so, if your overall usage on the day is higher than average, your relative saving will be considered to be higher (and vice versa). However, it could be the exact opposite of that and it really isn't clear at all.

I wouldn't expected it to be your relative use that gets compared for the in day adjustment - rather the general system demand or some sort of regional number.

Sort of "you usually use 2kWh, but everyone is using 50% more than normal today so we'll compare your saving against a 3kWh target".  This could end up with a stable use being counted as a saving - i.e. your usage didn't go up when everyone else's did.

Certainly could do with some clarification.

It very specifically is 'you', it's just not clear how the adjustment is calculated and what effect it has.

These half hourly averages are then subject to an ‘In Day Adjustment’ which will make adjustments based on how much energy you’ve used on the day of the session.

Then I think it's a poorly designed adjustment and ripe for gaming.

Petriix

[Deleted User] said:

Petriix said:

Deleted_User said:

Petriix said:

One thing which remains unclear is the notion of an 'In Day Adjustment' which they haven't clearly defined. The most obvious interpretation is that they will adjust the average of the previous 10 days' usage for that hour based on the relative use on the actual day so, if your overall usage on the day is higher than average, your relative saving will be considered to be higher (and vice versa). However, it could be the exact opposite of that and it really isn't clear at all.

I wouldn't expected it to be your relative use that gets compared for the in day adjustment - rather the general system demand or some sort of regional number.

Sort of "you usually use 2kWh, but everyone is using 50% more than normal today so we'll compare your saving against a 3kWh target".  This could end up with a stable use being counted as a saving - i.e. your usage didn't go up when everyone else's did.

Certainly could do with some clarification.

It very specifically is 'you', it's just not clear how the adjustment is calculated and what effect it has.

These half hourly averages are then subject to an ‘In Day Adjustment’ which will make adjustments based on how much energy you’ve used on the day of the session.

Then I think it's a poorly designed adjustment and ripe for gaming.

Potentially so, depending on how it actually works. But it might be designed to catch anomalous usage on previous days. I gather today's session is somewhat a 'trial run' so we'll hopefully have some clarity afterwards.

My average for this time slot is 593wH over the last 10 weekdays so there's a small potential for saving. I think I could get it under 100wH if I turn all but a few lights off. I can still use my laptop or Ipad on battery. I'd just be a bit gutted if my savings are adjusted down because my usage is lower than average over the whole day. I've done lots of washing and charged the car over the last couple of nights so expect to use a lot less in the off peak window tonight.

Petriix

Ok, I've found details about the 'In Day Adjustment' but, at first glance, it seems to disregard the entirety of the 10 day period supposedly used for the baseline.

Here's a link to the document

The bit I'm struggling to get my head around is that the formula as written: 

In Day Adjustment = ∑j (Unit Meter Point Metered Volume – Unadjusted Baseline Value)/ 6

where ∑j represents summation over the six Settlement Periods in the “In Day Reference Window”. The In Day Reference Window is defined as the three-hour period ending at one hour ahead of the first Settlement Period in Settlement Day D. BL01 Step 4 – Calculate Baseline Values Demand Flexibility Service | Procurement Rules For each ‘Domestic’ Unit Meter Point (and each Settlement Period within Settlement Day D), Baseline Value is calculated as:

Baseline Value = Unadjusted Baseline Value + In Day Adjustment 

...no matter what number you use as the Unadjusted Baseline Value, the result will always be the same. It adds it, then it takes it away so it's effectively removed from the result. That means the In Day Adjustment becomes the entire baseline. So the baseline is actually just the 6 half hour windows ending 1 hour before the trial on the same day.

That's clearly not the intention, but I can't see how you would use that formula to get a different result. Can anyone else figure it out?

Alnat1

If that's correct, are you saying we should use more electricity in the 3 hour period (that starts 4 hours before the trial begins) to gain the most reward?

So run the washer, dryer, dishwasher, cook in the oven, charge the battery storage a bit to cover trial time if needed? 

peter3hg

Petriix said:

Ok, I've found details about the 'In Day Adjustment' but, at first glance, it seems to disregard the entirety of the 10 day period supposedly used for the baseline.

Here's a link to the document

The bit I'm struggling to get my head around is that the formula as written: 

In Day Adjustment = ∑j (Unit Meter Point Metered Volume – Unadjusted Baseline Value)/ 6

where ∑j represents summation over the six Settlement Periods in the “In Day Reference Window”. The In Day Reference Window is defined as the three-hour period ending at one hour ahead of the first Settlement Period in Settlement Day D. BL01 Step 4 – Calculate Baseline Values Demand Flexibility Service | Procurement Rules For each ‘Domestic’ Unit Meter Point (and each Settlement Period within Settlement Day D), Baseline Value is calculated as:

Baseline Value = Unadjusted Baseline Value + In Day Adjustment 

...no matter what number you use as the Unadjusted Baseline Value, the result will always be the same. It adds it, then it takes it away so it's effectively removed from the result. That means the In Day Adjustment becomes the entire baseline. So the baseline is actually just the 6 half hour windows ending 1 hour before the trial on the same day.

That's clearly not the intention, but I can't see how you would use that formula to get a different result. Can anyone else figure it out?

The adjustment is calculated against the relevant settlement periods (half hour blocks) from the previous 10 days (4 days at weekend).
So if between 1pm and 4pm you used an average of 0.2kwh more per half hour compared to those relevant times from the previous 10 days, then your baseline for 5pm to 6pm would be adjusted upwards by 0.2kwh per half hour. Therefore it is still keeping your usage curve, but adjusted up or down based on your usage that day.

If you increased your usage in that 3 hour period you would gain more but, as it is averaged across 6 half hour slots, probably not worth creating extra usage unless it is just time shifting things you would need to do anyway.

[Deleted User]

Petriix said:

Ok, I've found details about the 'In Day Adjustment' but, at first glance, it seems to disregard the entirety of the 10 day period supposedly used for the baseline.

Here's a link to the document

The bit I'm struggling to get my head around is that the formula as written: 

In Day Adjustment = ∑j (Unit Meter Point Metered Volume – Unadjusted Baseline Value)/ 6

where ∑j represents summation over the six Settlement Periods in the “In Day Reference Window”. The In Day Reference Window is defined as the three-hour period ending at one hour ahead of the first Settlement Period in Settlement Day D. BL01 Step 4 – Calculate Baseline Values Demand Flexibility Service | Procurement Rules For each ‘Domestic’ Unit Meter Point (and each Settlement Period within Settlement Day D), Baseline Value is calculated as:

Baseline Value = Unadjusted Baseline Value + In Day Adjustment 

...no matter what number you use as the Unadjusted Baseline Value, the result will always be the same. It adds it, then it takes it away so it's effectively removed from the result. That means the In Day Adjustment becomes the entire baseline. So the baseline is actually just the 6 half hour windows ending 1 hour before the trial on the same day.

That's clearly not the intention, but I can't see how you would use that formula to get a different result. Can anyone else figure it out?

Unadjusted baseline is the 10 day bit.

In day adjustment is the amount that your in-day use (in the 3 hours ending 1 hour ahead of the trial) differs from this baseline.

So your actual target is your "normal" use, plus the extra that your in-day difference suggests it needs to have on.

Alnat1 said:

If that's correct, are you saying we should use more electricity in the 3 hour period (that starts 4 hours before the trial begins) to gain the most reward?

So run the washer, dryer, dishwasher, cook in the oven, charge the battery storage a bit to cover trial time if needed? 

Seems that way - but I'd like to hear other's reading of it.

[Deleted User]

from the previous 10 days (4 days at weekend).

Given that Octopus didn’t send out notifications of the session times until late the previous day, I am not sure how people could play the system - even if they wanted to. 

peter3hg

[Deleted User] said:

from the previous 10 days (4 days at weekend).

Given that Octopus didn’t send out notifications of the session times until late the previous day, I am not sure how people could play the system - even if they wanted to. 

Playing the system would be using more in the 3 hour period (starting 4 hours before the saving session) on the day of the saving session.


I've run my figures and I can see the effect of this adjustment.
My unadjusted average between 5-6 is 0.244kWh. However between 1-4 yesterday I used an average of 0.052kWh more in each half hour period than the previous 10 weekdays, due to the fact there was so little solar yesterday.
Therefore my adjusted average for 5-6 would be 0.348kWh.

I used 0.084kWh between 5-6 yesterday so I'm expecting the calculated figures to show I "saved" 0.264kWh.