Using net-metering and time-of-use tariffs with small scale embedded generation

April 1st, 2015, Published in Articles: Energize

 

The publication by NERSA of a study document on small embedded renewable generators covering a proposed simplified net-metering (NM) system [1] has been greeted with enthusiasm by the industry, giving a measure of certainty to the way forward. However, the announcement by the City of Johannesburg (CoJ) that it will introduce time-of-use metering on domestic accounts adds complications to the NM scenario, and may well affect the financial viability of NM schemes based on domestic rooftop solar systems.

The system proposed by NERSA would involve a kWh balance only, reconciled on an annual basis. Any surplus energy credit at the end of the annual reconciling is not carried over but forfeited to the supplier. The regulator’s document also states that the general aim is to discourage users from installing systems that provide all or almost all of their energy requirements.

Time-of-use (TOU) metering vs flat rate metering

Flat rate metering charges a fixed rate per kWh of energy used, irrespective of the time of day or season in which the energy is used. Inclining block tariffs (IBT) vary the rate according to the amount of energy used in a metering period, but are still not dependant on the time of usage. Flat rate provides an average charge over the day independent of what the price of generation is at that time.

The system proposed by NERSA does not take IBT systems into account. In this system a consumer may balance out energy which would have been charged at a higher block rate with energy produced, and can be seen to disadvantage consumers who are not on net metering.

Time of use metering sets a rate which is dependant on the time of use, and is designed to both recover higher generation cost during peak demand periods, and to encourage users to use less electricity during peak periods. TOU rates are lower than the flat rate for the off-peak period, and higher than the flat rate for peak periods, with a much higher rate for high peak periods.

TOU reflects the spot cost of generation. In off-peak cheaper baseload generators are running, and during peak demand period expensive peaking generators must be brought into service. TOU metering can have an impact on the effect of the net-metering system employed and could impact on the financial viability of an installation.

Flat rate metering

Consider how the kWh balance system proposed by Nersa would work. The system works on monthly meter readings with an annual reconciliation. The system could be implemented using existing meters or new smart meters, as it only needs the nett kWh consumed to operate.

TOU metering

If the straight kWh balance system is applied several scenarios could result depending on the TOU profile.

  • The consumer generates surplus energy (banks energy) during the off-peak period and consumes (withdraws energy) during the peak period. This disadvantages the supplier, and negates the aims of the TOU, as the consumer now offsets expensive units with cheap units banked previously. There is no incentive to reduce consumption during peak hours as credit is available from units banked during cheap hours. This scenario would apply to portion of the TOU profiles proposed for CoJ (Figs. 1 and 2).
  • The consumer generates during the peak hours and consumes during off peak. This is a disadvantage to the consumer as cheap units are being offset with units generated during an expensive time period. This scenario would apply to a TOU profile with its peak during the day and off-peak at night. Typical would be that applied in California [3], which has a peak rate from 12 h00 to 18h00 and off- peak outside those hours, so a domestic consumer would generate during this peak and consume during off peak.

To overcome these difficulties a system which assigns a value or weighting to kWh consumed or supplied to the network is necessary. One of the simpler common systems in use, and which is considered in the regulator’s proposal, is to apply the retail energy tariff charges for consumed units to the units supplied to the network. This means that units supplied to the network would accumulate a monetary value as well as an energy value, which would be used to offset the monetary value of units consumed. The system still has the feature that annual reconciliation would take place and any monetary credit would be forfeited to the supplier. This could cause problems and could result in contractual issues.

A second proposal would be to assign a weighting to kWh. For example a peak kWh would be equivalent to a multiple of the off peak kWh, e.g. in the CoJ example, a peak kWh in winter would be equal to 2,5 standard hours. This has the advantage of removing the monetary value and focusing of peak and off peak energy. In the first scenario above, the consumer would only receive a credit during the peak period of 0,4 kWh for each kWh generated during the standard period. In the second scenario this would be reversed.

However, assigning a weighting to kWh would also affect the annual reconciliation, as the surplus units accrued during peak periods could exceed consumption, and the units forfeited could then be significant, unless separate accounts are kept for peak and off peak periods which could complicate the system.

Whatever system is used, the metering systems would have to be considerably more complicated than the simple kWh balance proposed in the NERSA document.

Affects of TOU metering

Different values or weightings applied to NM can affect the overall economics of a system and may need much more complicated studies of consumption and generation patterns to achieve the optimum design for a domestic installation.

Time of use tariffs

To understand the effects we can consider the City of Johannesburg’s TOU 2014/2015 tariffs for single phase customers as an example [2]. Table 1 lists the TOU tariffs and Fig. 2 shows the summer tariff structure compared to the solar generation curve for a typical solar installation. Fig 3 shows the winter tariff compared to the solar generation profile.

Table 1: CoJ time of use tariffs
Tariff Summer Winter
Peak 109,89 c/kWh 262,09 c/kWh
Standard 86,93 c/kWh 104,65 c/kWh
Off-peak 68,39 c/kWh 73,38 c/kWh

 

Table 2: Time of use metering periods
Day Peak periods Standard period Off peak
Weekdays

07h00 – 10h00

18h00 – 20h00

06h00 – 07h00

10h00 – 18h00

20h00 – 22h00

22h00 – 06h00
Saturdays none

07h00 – 12h00

18h00 – 20h00

All other times
Sundays none none All hours

 

Fig 1

Fig.1: Summer TOU profile

 

Fig 2

 Fig. 2: Winter TOU profile

The effect of the TOU profile

The solar generation curve spans both the standard and peak morning generation periods. Any surplus generation during the morning peak would be credited at peak rates, and the balance at standard rates. The off- peak tariff period falls outside the solar generation window and the domestic peak consumption window and is not of concern. To achieve a balance between generation and consumption requires detailed information on both the consumption pattern and the generation pattern, taking into account variations in consumption and the TOU profiles over weekends, and the seasonal variation in generation between summer and winter.

During summer the potential for generating credits extends from 06h00 to 18h00 and during winter this contracts to 07h00 to 17h00. In addition the level of solar generation will be lower in winter and the peak period consumption is likely to be higher.

The generation period falls within the maximum of the solar curve and it is likely that the feed in to the grid would occur during this period. The peak rate in winter is 262,09 c/Kwh compared to 104,65 c/kWh for the standard rate. So all units generated during the standard period would be credited at 104,65 c/kWh. The units would be consumed during the evening peak period at a rate of 262,09 c/kWh, and during the off-peak at a rate of 73,38 c/kWh.

With this particular TOU it may be argued that the credits accumulated during the morning peak could balance the units consumed during the evening peak. But this would depend on the domestic load profile. For a flat profile the production schedule gives eight hours production at the standard rate and three hours at the peak rate. Remembering that the total production must be less than consumption, and all of the energy supplied to the grid must be balanced by energy withdrawn from the grid, it can be assumed that the full production is consumed during periods when solar is not producing. If the surplus energy during the morning peak is low, the energy generated in the standard period will have to make up the bulk of the energy consumed in the evening peak. Generally, the TOU system would result in a larger solar generator being required.

Compare this with the flat rate system where a straight comparison between average energy generated and average consumption can be used to size the system. The problem has been addressed by companies supplying solar power systems, and analysis programs based on hourly production and consumption figures over a full 8760 hours per year are available [3]. The critical issue here is accurate consumption data, as solar generation can easily be estimated.

Over sizing a system, with relation to generation during the morning peak, can result in excess credits which would be forfeited. Under-sizing could result in too big an imbalance between generation and peak consumption.

The storage option

A solution which could be considered to take advantage of the TOU tariffs would be to store all the surplus energy generated during the day and use this to offset the peak consumption units, effectively allowing units generated at the standard value to be used during the peak period without adjustment. For this to be effective the cost of storage would have to be less than the price differential between the standard rate and the peak rate, and would also depend on the actual peak and daytime consumption levels. Additionally, units generated over weekends during off-peak periods could be stored and used to offset units consumed at peak rates during the week.

References

[1] NERSA: “Small Scale Embedded Generation: Regulatory Rules”, www.nersa.org.za

[2] City Power: “Tariffs and charges 2014/2015”, www.citypower.co.za/Load%20Shedding/City%20power%20Tariffs%202014%20-%202015.pdf

[3] Sunlight Electric: “Net metering”, www.sunlightelectric.com/netmetering.php

Send your comments to: energize@ee.co.za

Related Articles

  • Evaluating the risks of South Africa’s energy sector to the country
  • Solar system replaces coal-fired plant
  • Considerations for the next generation of concentrating solar power systems
  • Onsite power for remote mining locations
  • Standby power generation set sizing