Martin Lee – October 2022
About the author
Founder member of OREF Martin Lee joined the Scottish Hydro Electric Board as a third engineer in Kirkwall in September 1986 and worked in Orkney until January 2001. After a period working in the south of England as a project manager and protection engineer, he spent the last 7 years of his employment with SSEN working in its future network section. This involved work across both the northern and southern parts of SSEN on projects looking at how electricity networks might be operated in the future.
The following is his personal recollection of the development of storage heating tariffs in Orkney. It may not be 100% accurate but it does tell an important story about a part of our history in which Martin was personally very involved.
White meter history
While much of GB used a common tariff for off-peak storage heaters called Economy 7, Scottish Hydro Electric had a longer off-peak period of 8 hours per day. Because the two rate meters used for the tariff had white cases to distinguish them from standard meters the tariff was called ‘the White Meter tariff’.
Switching between different rates used a mechanical time switch see figure 1. These time switches do not have any facility to change the time when moving between GMT in winter and BST in summer. The 8 hours cheap off-peak period was nominally from 11:30 pm to 7:30 am GMT in winter or from 12:30 am to 8:30 am BST in summer.
To avoid the peak load caused by switching on storage heaters, the mechanical time switches in Orkney had been set to different times. These ranged from starting as early as 10pm and as late as 1pm GMT but in all cases the cheap rate period lasted 8 hours. Often, as in the case of the time switch in figure 1, it appears to be set for 23:30 to 07:30 but the time on the clock is set to be an hour and a half fast so that it’s effectively operating from 22:00 to 06:00 in winter and 23:00 to 07:00 in summer.
However, this diversity of starting times still resulted in a peak load at around 1pm GMT with a rapid drop off in load shortly afterwards. To understand what was happening it’s important to understand how a 1980s vintage storage heater works.
There are two bimetallic strips within each heater, one controls the heat input and the second controls the heat output. While charging, a small heater heats up the output strip and closes the boost flap so that most of the heat being used to charge the heater is retained within the heater and does not quickly heat up the room at a time when it is not wanted, see figure 2.
After a while the input control strip switches the input off and as it cools it then closes again and allows charging to restart. The effect of this can be seen in figure 3.
Using my own house as an example the start time of the cheap rate electricity is 21:57 GMT slightly in advance of the 22:00 time it would have been set for in 1987. (not bad for a period 35 years!)
The blue line shows the actual consumption of the heater, it takes a steady load for around two hours and then starts to cycle off and on as the bimetallic control strip heats up and cools. As the heater stores more energy the on periods get shorter and the off periods get longer. The average of these periods is shown by the red line.
After about 5 hours the heater is close to achieving a balance between the heat input and the heat it gives out and the average line falls below a third of the peak load. The heater then continues to cycle but is effectively close to fully charged and in this instance takes under 30% of its peak load for the last two hours. While the White Meter tariff gave 8 hours instead of 7 hours cheap rate it can be seen that this would not result in an extra seventh in terms of the energy input into the storage heater.
System in operation
So, by 1987 Orkney had over a thousand homes with storage heating running in this way with start times varying from 22:00 to 01:00. The following graph (Figure 4) shows the addition of only 30 homes spread across the three-hour range, but it gives a good idea of the shape of the storage heater load would look like in blue, compared to everything switching on at 23:00 shown in orange. Some heaters would start cycling earlier than 2 hours into the charge and others would take a little longer. The red line shown in the earlier drawing had been adjusted, to show cycling starting as early as 1.5 hours and as late as 3 hours in both the blue and orange lines in figure 4.
This storage heater load was added to the underlying network load which showed a peak around 6pm and which steadily dropped off during the evening, returning to a steady base load by around midnight.
In figure 5 the Orkney total load profile in shown in blue. The storage heater part of the load is shown in grey and represents 23% of the daily load. This approximates to the winter peak load for the year. Though as noted elsewhere this is a recreation from a personal memory and not real data from 1986/87
Because the 33kV cable laid to Orkney in 1982 is effectively rated at 20,000kW, peak lopping would be required both around the tea time peak at 6pm and during the storage heater peak between 11:30pm and 2:00am. The cable capacity is shown in red.
Figure 6 shows the effect of not having the diversity in the starting time of the 8-hour cheap rate in terms of total load and storage heater load. The teatime and breakfast peak loads have not changed but the storage heater peak load has increased to 18,609kW and the overall peak load has increased to 26,087kW. The minimum amount of generation needed to meet the storage heater peak load would have been 10,800kWh from at least 6.1MW of generation, compared to only 3,600kWh from 2.7MW of generation with the diversified starting times. The duration of peak lopping for storage heater loads would have been about 2.5 hours. The saving of 7,200kWh represents a saving in diesel of about 1,800 litres on one of the coldest days. In 1986/87 peak lopping would only have been required on a few days of the year based on the diversified White Meter off-peak times but if they had not been introduced peak lopping would have been required on many more days.
The strategy of providing diversity in starting times for storage heaters worked well, but it was not enough. A better solution was required and the introduction of tele-switches (centrally controlled remote switches) allowed further diversity to be introduced into the operation of the White Meter Tariff.
The plan was to have two groups of customers each the same size with one group switching on at 10pm and then switching off again at midnight. The second group would switch on at midnight and then off again at 2am. The first group would switch on again at 2am and remain on for 6 hours switching off at 8am. The second group would switch on again at 4am and remain on for 6 hours until 10am. This scheme was introduced in 1989 in preparation for the winter of 1989/90. Around 500 customers were switched into each group. The nominal diversity gained is shown in figure 7
Both groups are only on together from 4am to 8am which corresponds to a low point in the load profile already existing for the diversified storage heater control.
There is no data about actual numbers at the time that tele-switches were brought into operation. There was a plan to change 1000 mechanical time switches to tele-switches with 500 in each group. But all new storage heater customers were also allocated to one of the groups of tele-switches. My recollection is that about 1,200 mechanical time switches were left from an original group of 2,200 and that the final number of tele-switches in each group was 800. This resulted in a load profile as follows. The estimated load of 84,000kWh per night across 2,200 homes gives an average figure per home per night of 38kWh. I have used this to scale back the original load profile and to create new ones for the two storage heater tele control groups. The underlying load has been increased causing peak lopping to be required more around teatime. These changes are shown in figure 8 in blue the peak load is around 22,900kWh. The grey line represents the load profile if the extra 600 new homes had been connected using mechanical tele-switches with a similar diversity to the original ones and no customers had been transferred to tele-switches. The areas between the grey and blue lines which are above and below the blue line match. However, the peak load has been reduced from 28,200kW to 22,900kW and the peak loping period has been stretched to 6.25 hours with the tele-switches. While peak lopping would still have been for 5.9 hours with the mechanical time switches. The minimum amount of peak loping generation required is 8,600 kWh and would have been 20,000 kWh with mechanical time switches. The savings in fuel per night are useful but the main aim was to reduce the peak load, this both reduces the amount of generation needed to meet it and made it easier to cope with the load following any failure of the cable from Thurso which would have required the power station at Kirkwall to run all the time.
Kirkwall power station had about 34MW of capacity but if the largest machine broke down it would only be capable of producing 27.5MW.
Total Heating with Total Control History
The load profile shown in figure 8 in blue was roughly the situation by the early 1990s. There were still times when the cable was not fully utilised between 10am and up to around 4:30pm, for a short period from 8:30pm to 10pm and again between 2:30am and 4am and between 4:30am and 7:30am.
Switching large blocks of load as shown in figure 8 at 22:00 and at 04:00 where changes of up to 7,000kW take place would not be much use in filling in these gaps. Faced with the potential requirement to install a second circuit to Orkney to cope with load growth and with adding additional generation capacity in Shetland, with either additional generation, or a second cable option being needed in the Western Isles within a few years, it was decided to use a relatively new tariff which could use the flexibility of tele-switches to fill in these gaps and make better use of the cable capacity in the case of Orkney and of the existing generation capacity in Shetland and the Western Isles.
This new tariff was called “Total Heating with Total Control”. It required that a minimum of 60% of the space heating requirements to come from storage heaters which Hydro Electric could control to manage the load. The remainder of less than 40% was typically panel heaters in bedrooms and a focal point fire in the lounge. All heating loads would be recorded by one meter while a second meter would record the rest of a household’s usage. It should be remembered that at this time Hydro Electric was operating as a vertically integrated company and extra costs incurred in supplying some energy at more expensive times was more than offset by the savings in reinforcement costs for additional cables and generator sets and additional fuel and maintenance costs for peak lopping in Orkney and the Western Isles.
Because of the requirement to provide a common tariff across the Scottish Hydro Electric supply area the Total Heating with Total Control tariff was available to any of Hydro Electric’s customers even though the need to use the tariff for load management was driven by avoiding very expensive reinforcements for the main island groups.
Orkney was allocated four tele-switch groups for domestic Total Control tariff and one for a commercial equivalent which was for storage heaters only. The Western Isles and Shetland had similar allocations while other areas only had 2 or 3 groups reflecting the fact that there was less value in managing the load in these other areas.
The tariff was designed on the basis that customers would get between 6 and 8 hours of input to their storage heaters each day with it being 8 hours in the winter. In practice the times set were left at 8 hours throughout the year in Orkney. There was also a minimum amount which had to be supplied overnight of 3 hours, during the evening one hour and the rest could be distributed around during the day or early morning in some cases. The tele-switches limited the number of on-off periods which could be used to 4 per day. I don’t know the final number of customers connected to THTC but by the time I left Orkney in January 2001 it was over 1200 with about 300 in each of the groups. THTC was offered to customers for many years and was still available in 2009 but I don’t know what happened in terms of total numbers of customers using the THTC tariff.
It was not possible to keep the peak loads down and as numbers of customers increased so did the peak loads However, the gaps below the cable capacity were filled up reasonably well by the use of THTC. Teatime peak load increased as well because of the panel heaters and focal point fires which were an integral part of the THTC tariff could be switched on. The effect of having 4,000 customers using one of the storage heating tariffs can be seen in figure 9. Note that the effect of load dropping off happens much sooner when a storage heater is energised after only a few hours since its last charge compared to the 16 hours it would experience with a traditional time switch. This resulted in the load profile looking like a saw tooth. While peak loads were able to be kept down the duration of peak loping required was extended up to 18 hours per day.
Figure 10 shows the thoretical loads which would have ocured by 1999 if THTC had not been introduced both with two tele-switch groups in blue and with the older diversified time switches in orange.
Summary of developments in storage heating loads
Note that the kWh in each day for 1998/99 do not totally match as I was not able to work out exactly what the effects of short bursts of charge to a storage heater were. The difference in 98-99 of around 7,000kWh per day across a total load of around 490,000kWh is about one seventieth of the total load, or across 1,200 THTC customers under 6kWh per day.
THTC continued to be installed long after 1999 but as I was no longer working in Orkney I don’t have a personal recollection of the numbers of customers involved.
In tabular form; figure 11 the ratio of peak load to average load can be seen to have been much improved by the use of THTC compared to continuing with either of the two White Meter options which shows a drop between 1986-87 and 1998-99. While White Meter with two tele-switch groups would have been a marginal improvement over the 1986-87 figures by 1998-99 in terms of the ratio between average daily load and peak load. However, the increase in peak load would have reduced by about 9.7MW which would have been a significant improvement. However, adding in THTC to the mix resulted in a much better improvement in the average to peak load ration and saved a further 4.1MW of peak load.
The effects on customers of THTC
Up to now the paper has shown how the THTC tariff came into being to limit network peak loading and to avoid very expensive network reinforcements primarily in Scotland’s islands. However, there was a cost to the use of this tariff in that substantial portions of energy supplied were outwith the traditional 10pm to 8am off peak period. The diversified times using mechanical time switches and the two groups using tele-switches also did this to a small extent. However, with THTC, more than half of the energy supplied was outwith the cheap off-peak period, though it did avoid the GB wide tea time peak period around 16:00 to 19:00. While Scottish Hydro Electric was a vertically integrated company it had been accepted that these costs were less than the benefits in not having to carry out the very expensive reinforcements which would be required to maintain both common access to tariffs across the Scottish Hydro Electric area and the provision of a heating tariff in areas off the gas grid.
I believe that somewhere between the merger of SHE and Southern Electric to form SSE and the sale of SSEs domestic supply business to OVO, the need for this cross subsidy got forgotten about. It may be that it was never accurately accounted for in the first place. The result is that the THTC when priced up against the costs of electricity on a half hour basis must be very expensive compared to a tariff where most of the energy supplied is between 10pm and 8am as is the case with the old White Meter tariff.
Currently under the Governments price guarantee from 1st October 2022 the standard default tariff rate for electricity in the North of Scotland is 33.08p per kWh. The tariff applied by OVO to a customer with a mechanical time switch giving 8 hours overnight cheap rate electricity is 18.44p per kWh off peak and 39.21p per kWh for the remaining 16 hours per day. The standing charge is virtually the same per day 51.16p for the white meter and 51.08p for the standard default tariff. A customer with a constant load 24 hours per day would be marginally better off on the White meter tariff than the standard default tariff.
THTC however in the same period is priced as follows. Heating load 25.87p per kWh other load 41.28p per kWh while the standing charge is the same as White Meter tariff at 51.16p per day.
Based on a daily consumption of 42kWh for heating and 8kWh for other uses the THTC tariff costs an additional £3.29 per day. This may be higher in winter and lower in summer.
The use of THTC may well be causing problems for many customers which are not due to a generally high price for electricity in the North of Scotland, but to the fact that they use electricity outwith the off-peak periods to heat their homes under the THTC tariff which is no longer cross subsidised by the distribution network operator.