Octopus Saving Session

milo_2020

£27 over 4 sessions, first 1 was experiment, second was a proof of concept (using 4 hour window), 3rd and 4th were actual using as much as i could during the 4 hour window, both days paid out £11+ and by using a house battery for the saving hour I 'saved' ~5kw during that hour.

pete-20-11

jessmist said:

I'm surprised we haven't had a saver session yet this week, especially with it being so cold.
Can't be much wind generating this week.

When a supplier signs up to the national grid demand flexibility test, they agree to run four tests aka saving sessions in the first 30 days and 2 tests per month thereafter, total of 12 tests.

maybe there will be one next week if the timings make sense as per above. 

Of course, they could decide to run a real one, rather than a test, although the last time the grid was thinking of doing that, they in the end decided it wasn’t needed. Quite why they changed their mind… who knows… but some have wondered if it was political interference! E.g. it doesn’t look good to say we are short on energy! 

OldUserNewName

razord said:

Here's my full calculation if helps people understand:

Session: 01/12/2022 17:00:00 - 18:00:00

[1] Session Period Averages for Qualifying Periods <- This is the average over the preceding 10 working days (excluding days which previously had a saving session) for each of the two 30m periods in the saving session.

17:00: 224.7Wh
17:30: 116.8Wh

[2] In Day Adjustment (Unadjusted Baseline) for Qualifying Periods <- This is the average over the preceding 10 working days (excluding days which previously had a saving session) for each of the 6 periods between E-4 and E-1 hours before the saving session.

13:00: 162.8Wh
13:30: 93.3Wh
14:00: 100.9Wh
14:30: 107.9Wh
15:00: 131Wh
15:30: 128.3Wh

[3] In Day Adjustment (Metered) for Session IDA Periods <- This is my actual usage during the 30m periods on the saving session day at E-4 to E-1 hours before the saving session.

13:00: 798Wh
13:30: 1547Wh
14:00: 573Wh
14:30: 1008Wh
15:00: 936Wh
15:30: 1410Wh

[4] IDA for Session IDA Periods <- This is my actual usage [3], minus the average from non-saving session days [2]

13:00: 635.2Wh
13:30: 1453.7Wh
14:00: 472.1Wh
14:30: 900.1Wh
15:00: 805Wh
15:30: 1281.7Wh

[5] In Day Adjustment on Baseline <- This is an average of the 6 saving session day periods at E-4 to E-1 before the event (Average of [4])

924.633333333Wh

[6] Complete Saving Session Calculation for Periods <- This is the numbers from [1], with the In Day Adjustment [5] added.

17:00: 1149.33333333Wh
17:30: 1041.43333333Wh

Savings, assuming 0w usage for Saving Session <- This is the monetary value for the saving session based on £2.25 per kWh saved.

17:00: £2.59p
17:30: £2.34p
Total: £4.93p (3943 octopoints)

Thank you for the above.

How close are people able to get to Octopus's own results?

I am having difficulty.  To date my calculations are 30% high for the first session; 6% low for the 2nd session; 6% high for the 3rd session and correct for the 4th session (not too difficult since my saving was negative, and so counts as zero).

As it may help others, I list below mistakes I was making, but have now corrected:

1. For no good reason, I was worrying about the difference between GMT and local time.  As, for me, Octopus have a complete set of recent smart meter data the earliest date of any relevance is 1st November (the earliest weekday contributing to the in-day adjustment for the 1st saving session, 15th November).  So for all relevant dates Local Time and GMT are one and the same.  I have now corrected my spreadsheet so that it consistently uses Local Time, as I believe that will be what is intended should these Saving Sessions go beyond the clock change next March.
2. Initially, I believed that the two half-hour periods of a saving session were totally independent.  In particular I believed that for the 1st session the in-day adjustment for the first half-hour was based on 13:00 - 16:00 and the in-day adjustment for the 2nd half-hour was based on 13;30 - 16:30.  It now appears that I was wrong and the in-day adjustment should have been based on 13:00 to 16:00 for both half-hours.

Points I'm not sure on:

1. Rounding, particularly when calculating averages.  If Octopus is performing any rounding, is it rounding up, rounding down or rounding to the nearest value - and when is it rounding and to how many decimal places?  This may make a slight difference to my results, but won't explain a 30% error!
2. Export: I have solar panels and for some relevant periods have a non-zero amount of export.  I have tried calculating my savings figures whilst totally ignoring the export, I have also tried modifying the import figures to net import figures (ie taking the import figure minus the corresponding export figure).  Both approaches give similar, but not identical, results. The Ofgem document specifying the procedure for calculation does refer to MPAN+ and MPAN- and states that for each half-hour period one of these should be zero.  I'm assuming that MPAN+ and MPAN- refer to import and export meters (in that order?), but don't understand why you can only have one or the other in any half-hour period.  It is quite common to have both some import and some export within the same half-hour period.

 Any observations welcome.

[Deleted User]

OldUserNewName said:

Initially, I believed that the two half-hour periods of a saving session were totally independent.  In particular I believed that for the 1st session the in-day adjustment for the first half-hour was based on 13:00 - 16:00 and the in-day adjustment for the 2nd half-hour was based on 13;30 - 16:30.  It now appears that I was wrong and the in-day adjustment should have been based on 13:00 to 16:00 for both half-hours.

Good you've fixed that - the in-day adjustment gets calculated once and is then static for the whole duration of the saving session that day.

OldUserNewName said:

The Ofgem document specifying the procedure for calculation does refer to MPAN+ and MPAN- and states that for each half-hour period one of these should be zero.  I'm assuming that MPAN+ and MPAN- refer to import and export meters (in that order?), but don't understand why you can only have one or the other in any half-hour period.  It is quite common to have both some import and some export within the same half-hour period.

It should be 'net energy transfer' in each period - so overall you either import or export (despite, as you note, actually having done both).  Although, of course, if you just added them up inside the calculation you would expect the same result.

paulypops

Looking like there will be a Session tomorrow DFS request for Test 5pm -7pm

paulypops

Saving session confirmed as 5:00pm until 7:00pm tomorrow you can opt in now.

tlcgrantham

paulypops said:

Looking like there will be a Session tomorrow DFS request for Test 5pm -7pm

Now confirmed thanks paulypops

[Deleted User]

No surprise - have you seen the wholesale market prices for that time tomorrow?!

razord

[Deleted User] said:

No surprise - have you seen the wholesale market prices for that time tomorrow?!

In theory the test events aren’t in any kind of reaction to wholesale prices are they? That’s only for live events?

[Deleted User]

razord said:

Deleted_User said:

No surprise - have you seen the wholesale market prices for that time tomorrow?!

In theory the test events aren’t in any kind of reaction to wholesale prices are they? That’s only for live events?

If you have to schedule a test - why not do it when the benefits are most useful?

Plus they can still continue to claim "we have not needed to activate this service", a useful political message to argue against indicators of shortage.