This cost-benefit analysis (CBA) was conducted by the Energy and Development Research Centre (EDRC) as part of a larger study on Environmentally sound energy-efficient low-cost housing study: Evaluation of performance and affordability of intervention technologies.

The consortium which conducted the overall study included the University of the Witwatersrand, University of Pretoria, Peer Africa and EDRC. The larger study is available in the EDRC Library (([email protected]). The client for the study was the inter-departmental Environmentally Sound Low Cost Housing Task Team, coordinated by Department of Housing and was funded by USAID

1. ENERGY SAVINGS AND COST INPUTS

These inputs were derived from the simulation process based on the Building Toolbox software as described in Chapter A-1. The inputs for the other interventions – lighting and water heating – come from a combination of the building simulation and other studies, as described below. Most of the interventions analysed in Chapter A-1 were related to thermal efficiency of the housing shell – in other words, the capacity of the building to control heating and cooling (regulate its internal temperature). North orientation and sunshading of north-facing windows in summer were not analysed separately since all the other interventions covered in A-1 assume north orientation of the house as a basic intervention.

0. Thermal improvements

Thermal simulations in Chapter A-1 of the wider study included a wide variety of interventions such as energy efficiency, water efficiency, and changes in municipal infrastructure usage. To analyse the costs and benefits of particular thermal improvements such as ceilings or exterior wall insulation, it was necessary to disaggregate these costs and the energy savings per specific intervention. This helped to identify not only the energy savings from a single intervention such as a ceiling, but also the savings from putting in a whole suite of interventions. The energy savings of each intervention are obviously not additive (three interventions that each save 50% of space heating energy use on their own do not add up to 150% savings), but the costs are generally additive. In addition, the costs of some interventions vary by region because different scales of an intervention are needed for different climates.

The tables below present the assumptions on incremental capital cost, energy savings and replacement costs, based on the outputs of the thermal simulation. Incremental costs refer to the capital cost of the intervention less any capital savings. For example, the installation of a solar heater nullifies the need for an electric geyser if the solar water heater has electrical backup. Note that here, as with all tables in the cost-benefit analysis chapters, the first 6 interventions refer to modifications to a standard 30m² RDP house. The next two refer to a 30m² RDP row house, where ‘shared wall’ shows only the costs and energy savings associated with moving from a free standing house to a row house design with two shared walls. ‘All SH Row’ would include a ceiling, roof insulation, wall insulation, proper window sizing, and interior partitions. ‘All SH Inf’ refers to modifications to a shack, which include a ceiling and exterior wall insulation.

Table 1: Incremental capital cost per intervention (Year 2000 Rands)

Intervention	Region			Comments
	U1 (CT)	U2 (G)	U3 (D)
Ceiling	957	957	957
Roof insulation	419	419	258	Thickness varied by climate
Partition	362	362	362
Wall insulation	736	1 474	418	Thickness varied by climate
Window	-593	-593	-593	Reduced total window glazing
All SH RDP	1 881	2 619	1 402	Includes all five previous interventions
Shared wall	-1 114	-1 114	-1 114	Reduced need for foundation and roof
All SH row	-105	-18	-380	Includes same as for standard RDP
All SH informal	1 247	1 247	1 247

Source: Simulation output of this study

Thermal simulation was conducted using a specific notional energy consumption (based on the climate and properties of the building). However, before one can apply this to the typical energy use patterns across regions, one needs to know the percentage of energy savings which can be applied as the typical value across the regions. The required percentages as derived from thermal simulations are presented in Table 2. Note that the combination of thermal improvement interventions were specifically designed to eliminate the need for space heating when used in combination – hence the 100% savings. This may well be overly optimistic, since the use of space heating holds both cultural and social meaning, and is not simply a basic economic and health necessity (Mehlwana & Qase 1999; Melhlwana 1999). This points to the need for long-term monitoring – both social and technical – of how households actually live in more thermally efficient homes.

Table 2: Energy savings per intervention (%)

Intervention	Region			Comments
	U1 (CT)	U2 (G)	U3 (D)
Ceiling	45	43	69
Roof insulation	5	8	12	Thickness varied by climate
Partition	7	8	12
Wall insulation	61	85	30	Thickness varied by climate
Window	6	11	9	Reduced total window glazing
All SH RDP	100	100	100	Includes all five previous
Shared wall	15	25	36	Reduced need for foundation and roof
All SH row	100%	100%	100%	Includes same as for standard RDP
All SH informal	100%	100%	100%

Source: Simulations output of this study

The thermal simulations and cost-benefit analyses assume that thermal efficiency interventions will last as long as the building itself (50 years), so that there is no need to replace them in the future. The exterior wall insulation and ceiling also provide important benefits on maintenance costs. Exterior insulation can reduce the costs of painting, and more importantly the need to repair cracks that would allow air infiltration. A ceiling reduces interior condensation, which in turn reduces rust and material wear and saves on maintenance. The magnitude of these savings, however, is not clear and, as with many other assumptions, need to be subject to proper field tests and monitoring. What is provided here is an estimate based on the simulations and estimates from Chapter A-1 of the wider study. In the absence of clearly disaggregated data, 50% of the annual savings has been apportioned to ceiling and 50% to the wall insulation in Table 3.

Table 3: Non-energy operating cost savings (R/year)

Ceiling	-93.5
Wall insulation	-93.5
All SH RDP	-187.0
All SH Row	-130.3
All SH Informal	Not applicable

Source: Simulations output of this study

1. Compact fluorescent lighting (CFL)

Although the initial cost of CFL is considerably higher than that of incandescent lamps, several studies have shown that the resultant energy savings outweigh the additional cost. Such studies include Clark (1997), Spalding-Fecher et al (1999), Praetorius and Spalding-Fecher (1998). The assumptions for this intervention are largely drawn from Spalding-Fecher et al (1999). The CFL that Eskom is promoting in its Efficient Lighting Initiative (ELI – the first large-scale energy efficiency programme in South Africa) has a bulb that can be separated from the ballast and is therefore cheaper to replace. Note that one incandescent bulb would only use 86 kWh per year (3.2 hrs/day x 365 days x 0.075 kW). Houses are assumed to have on average 4 bulbs. All of the costs per bulb and ballast are therefore multiplied by 4 for household net present value (NPV) calculations ( Chapter 3). In addition, the avoided expenditure on incandescent bulbs is counted as a benefit from installing CFLs.

Table 4: Lighting assumptions per bulb

	CFL	Incandescent	Comment
Initial cost (R/bulb)	R27*	R3.00	Bulb and ballast. Price indicated is subsidised price deemed acceptable to customers
Bulb life (hours of use)	8 000	1 000
Ballast life (hours of use)	40 000	N/a
Power rating (W)	19	75*	75% energy and demand savings
Hours of use (hours/day)	3.2	3.2
Bulb life (years)	8	0.86	Based on useful life and usage
Ballast life (years)	34		‘
No. replacements (bulb)	6		Over 50 year life of building
No. replacements (ballast)	1		‘
Replacement cost (R/bulb)	13
Replacement cost (R/ballast)	30

* A 75W bulb here represents a mix of 60W and 100W bulbs

Source: CBA output of this study

 

2. Solar water heating (DSWH)

One critical question related to replacing electric storage geysers with DSWH is whether some type of back-up energy source is required to guarantee hot water on demand on cloudy days or when more people are using the home’s facilities. Whether ‘enough’ hot water is available at all times depends on the weather, the size of the collector and storage tank, but also importantly on the lifestyles of the residents. While there are examples of homes that have solar water heating in South Africa with no backup (Holm 2000), other analysts and consultants involved in providing DSWH to low income communities point out that often only 60-70% of the energy needed (and hence hot water) can be provided by solar energy, and so some back-up is necessary to guarantee hot water on demand (Morris 2000; Thorne et al 2000). The study assumes that some backup is needed, and the energy savings are 60%. While the cost savings will vary if the solar water heater backup is non-electric, additional information is needed to quantify these costs.

The up front costs for a 100l indirect DSWH with a 1.8 m² collector are taken to be R4 000, while the avoided costs of not having to install an electric geyser are R2 200 (see Chapter A-1). This would provide enough hot water for a typical family of 6. Greater use would increase the use of electrical back-up and so reduce the energy savings. However, it should be noted that the usage patterns for piped hot water in low-income areas (and especially newly electrified ones) is not well understood (Morris 2000).

DSWH are relatively durable, with entire systems lasting for 15 years or more. The tank element for electrical backup might need to be replaced before then. There is limited experience with DSWH over longer time periods. The study assumes that the replacement cost will be about half of the initial cost because of greater efficiency of production in the future and the fact that not all of the system components would need to be replaced.

Table 5: DSWH assumptions

	SWH	Elec storage	Comment
Initial cost (R)	R4 000	R2 200	Includes cost of back-up
Life (years)	15	15
Energy savings	60%
No. replacements	2		Over 50 year life of building
Replacement cost (R)	R2 000

Source: Simulations output of this study

 

3. Solar home systems

An off-grid solar home system (SHS) will provide electricity for three or four lights, a black-and-white television set and/or a radio/hi-fi. Low-power DC appliances (such as a fan or sewing machine) can be powered by medium-sized systems, but these are uncommon in South Africa. With the addition of an inverter (R300 to R 1 500), small AC appliances (colour television, AC sewing machine) can be operated. While refrigeration is possible, it is relatively expensive. LPG- or gas-powered refrigeration is usually more economical for domestic or retail applications (Banks 1998A). Off-grid electrification is not normally used for cooking, space or water heating as it is too expensive to generate and store the large amounts of energy required.

A medium sized system (50W, 100 Ah) has a total capital cost of R2 435, and lifetime maintenance/replacement costs of R2 230 (excluding bulbs) (Banks 1998a). This represents R187 per year in maintenance and replacement costs. SHS is most desirable if it completely replaces the grid connection so that one avoids the cost of readiboard, prepayment meter, and distribution infrastructure. Chapter A-1 of the wider study presents an estimate of R580 and R852 for the costs of a readiboard and prepayment meter respectively. This leaves a net capital cost of approximately R1 000. The costs of connecting new households to the grid varies between R1 300 and R1 800 per household in urban and peri-urban areas (Banks 1998b). To deduct these savings from the capital cost as well, however, would make the net capital cost negative – implying that all urban electrification should have been done with solar home systems. Why is this?

The major difficulty here is that, because SHS are considered as an intervention to save energy used for electric lighting and light appliances. It does not satisfactorily replace other end-uses such as cooking which might utilise grid electricity. A proper analysis of going ‘off the grid’ must include all major end-uses and how they would be provided with and without the grid. There could be other avoided costs such as coal used for space heating. An ‘energisation’ package would also include new costs such as access to LPG appliances and the higher cost of LPG per MJ. The negative NPV of SHS in the CBA arises from considering only some of the benefits but counting all direct costs. It should not be taken as an indication that SHS in semi-autonomous urban houses are not viable.

Off-grid electricity has not been considered for urban low-cost housing in South Africa to date. Consumers still show a desire for the option to use high-demand electrical appliances even if they do not use electricity for all end-uses (not to mention the necessity of high electricity usage to pay off the utility and government’s investment in the connection). In view of the above, the costs of the readiboard and meter have been deducted, but not those of the grid connection even though in theory these two costs are linked to one another.

There are three options to refine this analysis. One would be not to exclude any of the ‘avoided’ distribution infrastructure costs – and so assume that the house still has grid electricity for some end-uses. This would be fairly simple analysis, but it would make the intervention even more expensive and unrealistic. The second option would be not to include the SHS in this study at all – because there is no adequate data on the whole range of energy-uses by households.

Opportunities to explore the full benefits of these systems in urban areas should be explored. One possibility would be to do a complete analysis of an urban ‘energisation’ package – across all end uses, including any additional operating costs, and ensuring an equivalent level of service as grid electricity. This would be very useful in terms of understanding the barriers to renewable energy systems and the implicit costs of a political decision for grid electrification. The most useful analysis at this stage would seek to address the question whether the additional costs of an urban ‘off grid’ home could be offset at all by climate change-linked funding (see Chapters 3 and C-2 in the wider study which covers information gaps for further research)

2. FUEL-USE PATTERNS IN URBAN SOUTH AFRICA

Data on cost savings on low-cost housing is available from a number of sources. Simmonds and Mammon (1996) is a key source for the data used in the CBA process of the study. The study collated findings of several other studies on energy services in low-income urban households. constitutes another significant data source used in the CBA. All the data from Simmonds and Mammon (1996) was derived from various sources including the 1993 country-wide survey by South African Labour and Development Research Unit (SALDRU) on living standards in South Africa. Data from the National Electricity Regulator (NER 1998) was important for understanding the proportions of electrified and non-electrified households per province.

The fuels considered in this study were electricity, paraffin, wood, coal and gas. Other fuels which were not considered were candles, gensets (petrol and diesel) and lead-acid batteries. End uses not covered include cooking and refrigeration, as these do not relate directly to the housing shell design.

There is no current data set that combines fuels by end-use, by income group and by province. Although the energy-related data from SA Focus is the most recent data available, it expresses energy consumption as an expenditure across all end-uses and not by end-use. This could not be used for analysing particular end-uses.

The analysis of fuel-use patterns combined information on household consumption by fuel-type with the share of homes using particular fuels. Simmonds and Mammon (1996) show household consumption profiles by fuel-type and region, represented by Cape Town, Durban and Johannesburg. They also show percentage estimates of how much fuel is used per particular end-use. Consumption figures for each end-use were then obtained by multiplying the consumption of each fuel in homes that used it with the percentage of homes using that fuel for a particular end-use. These monthly figures were converted into annual consumption, assuming that households need space heating for four months in a year. The study assumes that lighting and water heating is required throughout the year.

The fuel-use patterns and percentage share of households using particular fuels for different end-uses are shown in the tables below, with space heating in Tables 6 and 7.

Table 6: Annual consumption for space heating by region and fuel

	U1 (CT)	U2 (Jhb)	U3 (Dbn)	Units
Electricity	387.8	358.4	387.1	KWh
Coal	371.7	743.4	247.8	Kg
Wood	0	0	0	Kg
Paraffin	49.2	21.0	22.8	Litre
Gas	6.9	2.0	2.7	Kg

Source: CBA output of this study based on Simmonds and Mammon (1996: 70, 73-6)

Table 7: Share of houses using fuel for space heating by province

	U1 (CT) (%)	U2 (Jhb) (%)	U3 (Dbn) (%)
Electricity	75	69	54
Coal	2	5	3
Wood	0	0	0
Paraffin	19	23	38
Gas	2	1	0

Source: CBA output of this study based on Simmonds and Mammon
(1996: 70, 73-6) and NER (1999: 16)

Given that coal is primarily available inexpensively in Gauteng and Mpumalanga, and the climate is considerably colder, it is understandable that the coal consumption figures are highest for this region. Both Cape Town and Durban have higher levels of paraffin usage than Johannesburg. The low percentage of homes using coal for space heating in Johannesburg is, however, surprising. In a review of the SADLRU (1993) survey, Simmond and Mammon (1996) observe that the survey focused more on established households. Such households are more likely to use proportionately more electricity. In addition, the study considered households living in formal housing and not in shacks. In many cases the move from informal to formal housing also stimulates additional electricity use, although this process is by no means comprehensively understood. Finally, electrification levels are highest in Cape Town, which also explains the higher use of electricity in those households.

Obviously, no households use coal or wood for lighting. One can note that the electricity consumption for lighting does not vary significantly across regions, but paraffin consumption does. Durban has a lower share of homes using electricity and paraffin for lighting (54% and 9% of total households respectively). A closer observation shows that the remaining percentage of households use candles for lighting, however candles are not included in this cost-benefit analysis.

The difference in paraffin prices between regions is due to different fuel use patterns between regions, with Gauteng and Kwa Zulu Natal using less than Cape Town (Simmonds & Mammon 1996). By contrast, coal-use in Gauteng is higher than for other regions. Paraffin is cheaper at the coast than inland and that is possibly why less of it would be used for lighting in Gauteng. The use of alternative sources of lighting not included in the study, notably candles, further explains the variation.

Table 8: Consumption on lighting by province and fuel

	U1 (CT)	U2 (Jhb)	U3 (Dbn)	Units
Electricity	332.4	307.2	331.8	KWh
Coal	-	-	-	-
Wood	-	-	-	-
Paraffin	123	52.5	57	litres
Gas	-	-	-	-

Source: CBA output of this study based on Simmonds and Mammon (1996: 73-4)

Table 9: Share of houses using fuel for lighting by province

Baseline	U1 (CT) (%)	U2 (Jhb) (%)	U3 (Dbn) (%)
Electricity	80	72	54
Coal	0	0	0
Wood	0	0	0
Paraffin	16	6	9
Gas	0.4	0.3	0

Source: CBA output of this study based on Simmonds and Mammon
(1996: 73-4) and NER (1999: 16)

Tables 10 and 11 present water heating energy use and share of households using particular fuels. The consumption figures on water heating are high. They are based on low consumption level averages rather than consumption in well-established homes . Low monthly consumption for all end-uses is 345 kWh per month. Annualised and multiplied by 40% of consumption used for water heating (Simmonds & Mammon 1996: Tables 5.9 and 5.5).

Table 10: Consumption on water heating by province and fuel (kWh)

	U1 (CT)	U2 (Jhb)	U3 (Dbn)	Units
Electricity	1.656	1.656	1.656	KWh

Source: CBA output of this study based on Simmonds and Mammon (1996: 74-6)

Table 11: Share of houses using fuel for water heating by province

	U1 (CT) (%)	U2 (Jhb) (%)	U3 (Dbn) (%)
Electricity	74	68	31
Coal	0	5	4
Wood	3	1	28
Paraffin	17	23	19
Gas	5	1	2

Source: CBA output of this study based on Simmonds and Mammon (1996: 44),
Afrane-Okese (1998: 119) and NER (1999: 16)

Overall, the patterns of fuel use show that electricity is commonly used for space heating and lighting and water heating but not exclusively. In the calculations for the cost-benefit analysis, only electricity is considered for water heating and lighting – because the interventions deal with alternative electricity supply. All fuels are included in space heating component of CBA.

3. FUEL PRICES

DME (1999) contained valuable price data for paraffin, gas and coal. For all the fuel prices from the DME, the coastal price was taken as the price in Durban and Cape Town and the Reef price for Johannesburg. In the absence of separate rural prices, we have assumed a price higher than the Reef price for rural areas. Due to the South African system of determining fuel prices by the global crude oil market, the figures for paraffin may have changed this year with the major swings in oil prices even though the government attempts to minimise the volatility of local prices. Table 14 shows both the coastal and the Reef prices of LPG. Retail prices were used in all cases since the CBA focuses on the economics at household level.

Coal price data were derived from two sources: Simmonds and Mammon (1996) and DME (1999). The first source cover Cape Town and Johannesburg and shows the price in Johannesburg to be about half the price in Cape Town. This is possibly because coal is more readily available inland. The price of coal in Durban was assumed to be similar to that in Cape Town. Since no price was available for the rural areas, the current DME price was used (DME 1999). Table 15 shows the retail price for domestic coal.

Simmonds and Mammon (1996) provide retail coal-prices for Johannesburg and Cape Town as well as prices for wood. Since wood price data for the Cape Town region were not available, the average of Durban and Johannesburg prices was used.

1. EXTERNAL COSTS OF ENERGY USE

The external costs of energy supply reflect the environmental and other social costs associated with their use. These costs can be especially difficult to quantify in monetary terms and are usually expressed as ranges rather than precise figures. Previous research on external costs of energy supply in South Africa relates to environmental costs of electricity generation, costs of fires and burns associated with paraffin use in the home, and the costs of illness and death caused by indoor air pollution from coal and wood burning (van Horen 1996a and 1996b). This analysis distinguishes between the global external costs associated with greenhouse gases and the local environmental impacts that reflect immediate health impacts, for example, from indoor air pollution. The total external cost is applied in the derivation of social benefits, while the cost of local impacts is applied in the derivation of the cost of avoided greenhouse gas emissions. The results are presented in Chapter 3 with and without external costs to illustrate the impact they have on the results.

Local external costs are taken from van Horen’s study of household external impacts and impacts of electricity generation (van Horen 1996a). The damage cost of greenhouse gases is based on the work of Fankhauser and Pearce for the IPCC and reported in Pearce (1995). The study recommends a damage cost of US$22 per ton of carbon, or US$6 per ton of CO₂. This translates to R37/t at 6.2 R/US$ (reported in Pearce 1995). The external cost assumptions are summarised in Table 18. For more detail on the calculations see Spalding-Fecher et al (1999).

Table 18: External cost assumptions by fuel (1999 Rands)

Fuels (units)	Local impacts		Greenhouse gas impacts		Total external cost
	R/GJ	R/unit	R/GJ	R/unit	R/GJ	R/unit
Electricity (kWh)	2.6	0.01	10.7	0.04	13.3	0.05
Coal (kg)	4.7	0.13	3.9	0.10	8.6	0.23
Wood (kg)	25.7	0.40	0	0	25.7	0.40
Paraffin (litre)	53.6	2.04	2.7	0.10	56.3	2.14
Gas (kg)	-*	-*	2.1	0.10	2.1	0.10

* No research available on local impacts of LPG

Source: Spalding-Fecher et al (1999: 35); van Horen (1996a: 170); IPCC (1996) and Pearce (1995)

2. HOUSING STOCK AND BACKLOG

Low-cost housing was interpreted as costing between R7 500 – R 17 250 and fully funded by government subsidy (. Housing data were drawn primarily from whose data are based on primary research with developers as well secondary sources such as . Additional information on housing backlog was obtained from the Department of Housing (DoH, 2000). The data for low-cost houses built between 1 April 1997 and 31 March 2000 was available on a provincial basis.

National estimates data were available for 1994-97. The data wwas converted to provincial figures using the same percentage breakdown as the period 1997-2000. The data on the number of houses built between 1994-97 varies considerably between the two sources cited above. Hendler (2000) gives a total of 450 876 houses while DoH (1999) estimates show 86 700 units. Assuming that that the real number lies somewhere between the two, a simple average was taken for this analysis. This is logical given that it is lower than the houses completed between 1997 and 2000, when the mass housing drive accelerated. In addition, the 1997-2000 data has the highest level of accuracy as it is based on primary research with developers as well as secondary sources. This total number of houses was then allocated to each province using the same breakdown as for the 1997-2000 period.

Totals for houses built in urban areas from 1960 to 1977 were obtained, and were similarly apportioned by province. Very few low-cost houses were built during the 1977-94 period. Consequently, no additions to housing stock from this period have been included. The urban fraction for 1994-2000 is assumed to be the same as the share of subsidy allocation to urban areas in the last few years (Hendler 2000). Totals for the existing housing stock were calculated by adding the urban fraction of housing stock for the periods 1960-77, 1994-97 and 1997-2000 for each province.

Totals of housing backlogs per province were provided in Hendler (2000) which draws on data from the Department of Housing as quoted in The total number of houses that needed to be built as of June 1998 was 2 603 717. These data are comparable with the latest estimate by the DoH of 2 784 193 for June 2000 (Bosch 2000).

To estimate the proportion of the housing backlog in the urban areas, the study assumed that three-quarters of the housing would still be built in urban areas and only one quarter in rural areas. The Department of Housing and the Energy Development Group were consulted for their views, but no accurate percentage breakdown could be obtained. Given the new commitment to a rural bias in government, this breakdown might change in future. However, this change is recent, and much housing in rural areas is not provided by government. The most recent data available on informal houses was from the 1996 Census .

In order to aggregate findings for households in the three regions, the study assumed that these regions are proxies for parts of the whole country. The study therefore combined the provincial data for housing stock, backlog and informal houses for different provinces as follows: Region U1 combines the Western, Northern and Eastern Cape; U2 is Gauteng and Mpumalanga; and U3 is KwaZulu Natal, Free State, North West and Northern Provinces. The consolidated figures are shown in the Table 19.

Table 19: Number of houses in target group for each intervention per region (000s)

	RDP 30 m2 house – space heating						Row house – space heating		Informal house	Lighting		Water heating
	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
U1 (CT)	658	658	430	658	430	430	430	430	334	658	430	430
U2 (Jhb)	916	916	709	916	709	709	709	709	562	916	709	709
U3 (Dbn)	1 078	1 078	812	1 078	812	812	812	812	555	1 078	812	812

Source: CBA output of this study

Some interventions can be applied to existing housing stock and new houses (i.e. the backlog), while others (e.g. window sizing and partitions) are only applicable to new houses. Thermal improvements in informal housing are taken as one comprehensive intervention.

In order to aggregate results from the household to national level, the number of houses in the target group for each intervention needed to be defined. Only newly constructed housing (based on housing backlog data) was taken as the target group for the partition, window size, all packages of thermal improvements in all formal houses, the shared wall in the row house, the SHS and the DSWH. For the following interventions, both existing housing stock and new houses constituted the target group for ceiling, roof insulation, wall insulation and compact fluorescent lights. For the thermal interventions in informal housing, the number of informal houses was used as the target figure.

3. OTHER GENERAL ASSUMPTIONS
1. Discount rates – consumer and social

Discount rates reflect the time value of money and are used in the calculation of the present value of future costs and revenues. Discount rates depend on the specific perspective adopted. In this context there are three different perspectives of interest.

Firstly, there is the national perspective which addresses the question as to whether an intervention is in the national or economic interest. The relevant discount rate here is the social discount rate, which reflects the return which can be expected from a similar category of investment. It is usual for national authorities to indicate a discount rate which is used for the evaluation of a wide range of cross-sectoral projects. In developing countries, where capital is considered scarce, high discount rates are usually used. This is the case in South Africa where a real discount rate of 8% is usually recommended for government infrastructure projects. This rate essentially reflects the cost of capital available to government.

Consumer discount rates are expected to be significantly higher as low-income households generally pay a premium for capital. In fact, many low-income households rely on especially punitive sources of capital such as hire-purchase and so-called ‘loan sharks’ (see Banks 1999). While this is partly a reflection of the high transaction costs of dealing with small amounts of capital usually loaned to households, it is principally a consequence of the lender’s evaluation of the risks of lending to these households. Nevertheless, it is probable that the credit market for low-income households is an extreme example of a case where the usual market systems fail to provide an appropriate response to the needs of this group of households (imperfect market). It is difficult to determine an appropriate discount rate for this group, partly because circumstances can change significantly for different individuals. Nevertheless, a discount rate of 30-40% seems appropriate, and 30% is chosen for the CBA.

2. Inflation/choice of currency year

All currency values are in constant 1999 Rands. Where figures (e.g. fuel prices) were given in Rands for other years, the appropriate conversion factors were used to reflect inflation. The inflation factors, based on the gross value added deflator, are shown in Table 20.

Table 20: Adjustment factors for conversion to 1999 prices

Year	Inflation rate	Deflator (multiply by this factor to get 1999 prices)
1988
1989	16.1	2.8
1990	16.3	2.4
1991	16.44	2.1
1992	14.5	1.8
1993	12.0	1.6
1994	9.3	1.5
1995	10.3	1.3
1996	8.0	1.24
1997	8.0	1.15
1998	7.6	1.06
1999	6.5	1

Source: Adapted from SA Reserve Bank (1999)

 

4. RECOMMENDATIONS

1. POTENTIAL CONTRIBUTION TO SOUTH AFRICA’S CLIMATE-CHANGE AGENDA
1. INTRODUCTION AND METHODOLOGY

The objective of this chapter is to quantitatively assess the contribution of the specific interventions in the CBA to the mitigation of greenhouse gas emission. While there is some overlap with the more general discussion in Chapter C-2 (covering information gaps for further study in the wider study), this chapter focuses only on the interventions included in the CBA, their impact on annual emissions of carbon dioxide in South Africa as well as their cost effectiveness.

The scope of this analysis covers the carbon dioxide emissions from the combustion of fossil fuels both in homes and at electricity generation stations. In addition, the coal bed methane emissions associated with coal mining and coal-fired power production have also been included in the emissions factor for electricity (see Praetorius and Spalding-Fecher 1998). This study only covers direct emissions from combustion. ‘Embodied emissions’ (emissions that are a product of embodied energy) arising from production, distribution and site-related operations for building materials used in low cost homes are not included.

As with the economic benefit calculations presented in Chapter 3 the national totals were built up from household level analysis. Avoided emissions per household were based on avoided energy-use and South Africa-specific emissions factors per delivered unit of energy as shown in Table 21.

Table 21: CO₂ emission factors for South Africa compared to IPCC default factors

	South Africa		IPCC default factors*
Energy source	kg CO2 /GJ	Kg CO2/unit	kg CO2 /GJ
Coal**	104	2.81/kg	94.6
Paraffin	71.5	2.72/litre	71.3
Gas	56.1	2.79/kg	56.1
Electricity delivered**	287.4	1.03/kWh	N/a
Wood	0	0	n/a

Source: Davis & Horvei (1995); IPCC (1996); Praetorius & Spalding-Fecher (1998: 37)

* IPCC default CO₂ emission factors for comparison.
** Includes CO₂ equivalent for methane emissions related to coal mining.
*** Excludes losses in transmission and distribution (T&D).

The emission factor for electricity was calculated based on the coal burned in Eskom’s power stations for electricity generation in 1996. In that year, Eskom burned 85.4milion tons (mt) of coal which led to an emission of 157mt CO₂ or 266 kg CO₂/GJ electricity generated in coal-fired power stations. When applied to the total electricity generated by Eskom, this amounts to 243 kg CO₂/GJ. A 10% loss is assumed for transmission and distribution (T&D).

The CO₂ equivalent of methane coming from mining coal for electricity generation was also included. This is approximately 473 000t of methane or approximately 9.9mt of CO₂ equivalent. Altogether, the emission factor used in this report amounts to 287.4 kg CO₂/GJ electricity delivered to the end-user. For more detail on this calculation, see Praetorius and Spalding-Fecher (1998).

Avoided GHG emissions are calculated for each household by multiplying the annual energy savings (unit energy / year) by the emissions factor (kg CO₂/unit energy) for each fuel. Avoided emissions by fuel and region (tonnes CO₂/year) is derived by aggregating household values by the number of households using the relevant fuels in the different regions. These are then aggregated across regions for each fuel (tonnes CO₂ / year). Finally, a national figure of GHG emissions avoided is calculated by combining the results from all fuels (tonnes CO₂/year). (Chapters 3 and 4 provide more details on the aggregation process).

2. GHG-EMISSION SAVINGS AND COSTS BY INTERVENTION

Figure 1 shows the total annual avoided emissions for each intervention. Not surprisingly, the interventions that save the most energy (e.g. ceilings, wall insulation, solar water heating) also save the greatest amount of emissions.



Figure 1: Avoided emissions (t CO₂ /yr) for each intervention nationally
Source: CBA output of this study

The avoided emissions should be understood in the context of total emissions of 373mt of CO₂-equivalent (mt(CO₂)) in 1990, with the energy sector contributing 89% of the total (van der Merwe and Scholes 1998). In 1990, the residential sector contributed approximately 10mt of CO₂-equivalent (De Villiers & Matibe 2000). The interventions analysed here show potential GHG reduction ranging between 0.05 and 0.6 mt(CO₂) per year which is 0.1%-6% of the residential total. While this is a small amount compared to national emission level, it is significant considering that the study only covers low-income households and not all households. Perhaps more importantly, at an international market price of $10/t of CO₂ (65 R/t), 3mt of CO₂ is worth roughly R200 million per year – a substantial contribution to local development projects.

Having calculated a national total of avoided emissions, the cost of avoiding emissions (or cost of mitigation) can be derived. This is done by dividing annualised social NPV for the intervention by its annual avoided emissions (t CO₂) to give a cost per tonne (R/t CO₂). To avoid double counting of economic benefits, we have only included avoided external cost in the NPV – not global. Table 22 shows the resulting costs in Rands.

Table 22: Cost of avoided emissions per tonne of CO₂ for each intervention

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
Including local externalities	-391	732	679	-393	-1 879	-375	-362	-730	635	-220	450	-6
No externalities	-327	796	743	-336	-1 819	-314	-350	-668	696	-211	459	3

Source: CBA output of this study

The interventions that are already cost-effective before considering the global environmental benefits have a negative cost of mitigation – in other words, reducing emissions saves money rather than increasing costs to the economy. These are often called ‘no regrets’ or ‘win-win’ climate change mitigation projects. Clearly many of the energy efficiency interventions are ‘no regrets’ options. The higher costs for roof insulation, partitions, informal houses and SHS are discussed in detail in Chapter 3, and relate to relationship between additional capital costs and incremental energy savings as well as the services provided.

The global attractiveness of intervention projects as mitigation options within mechanisms such as the Clean Development Mechanism (CDM) and the Global Environmental Facility (GEF) depends on how they compare to global estimates of marginal abatement costs for CO₂. No industrialised-country investor would invest in a South African project if the cost per tonne of carbon were higher than they would have paid in their own country or on an international ‘spot market’ for carbon credits. A number of studies have tried to estimate this ‘market price’ for carbon, and have been collated by Vrolijk (1999). Mitigation costs per tonne of CO₂ are projected to be in the range of US$13-$42/tC or $3.5 to $11.5/tCO₂ . For comparison, the values above are also presented in US$, assuming an exchange rate of R6.80 / $1.00.

Table 23: Cost (US$/t CO₂ ) of avoided emissions for each intervention.

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
Including local externalities	-58	108	100	-58	-276	-55	-53	-107	93	-32	66	-1
No externalities	-48	117	109	-49	-268	-46	-51	-98	102	-31	67	0

Source: CBA output of this study

In the context of international negotiations on CDM and the increasing interest in South Africa as a host for projects, it is important to remember that the cost of a project for a CDM investor is not the same as the cost to the country. One of the major differences in analysing CDM projects versus GHG mitigation options is that the financing arrangements do not necessarily coincide with the changes in economic costs. In other words, in ‘traditional’ mitigation analysis one would basically compare the life-cycle costs and GHG emissions with and without the intervention while taking into account as much as possible, full economic costs, appropriate (social) discount rates among other factors. One would then compare these two to get the incremental life-cycle cost which, if divided by the total emissions savings, gives a measure of cost effectiveness. The GEF operational guidelines refer to this increment as the market barrier removal (reduction) costs.

CDM investors are not interested in market barriers or the future local savings per se. Instead, they would be interested in realising profits out of such investment. This would mainly depend how many credits they can generate per dollar invested in the project. This may not bear much resemblance to the cost per ton identified in the mitigation study for a variety of reasons (e.g. rules on sharing of credits, financing mechanisms, equity partnerships, etc). Analysing how CDM investors might see these particular interventions is beyond the scope of this study. However, as part of the information gaps identified in this study, Chapter C-2 provides a discussion of the necessary policy and legal framework required for South Africa to become a competitive player in this field for low-cost housing market. Thorne et al (2000) provides further details on the international dynamics of this market.

3. CONCLUSIONS

The interventions analysed here make a contribution to avoiding greenhouse gas emissions, which is substantial when compared to total residential greenhouse gas emissions even if they are small compared to national totals. The selling point for these mitigation options, however, is their low cost of mitigation and their significant development benefits – in terms of reduced local household expenditure, improved health and potentially increased employment. These benefits are explored in more details in Part B of the wider study which deals with indirect cost savings. The low GHG-related cost-potential of the interventions combined with their local socio-economic benefits, make them attractive for international climate-change linked investment that would support national development priorities.

2. COST-BENEFIT ANAYSES OF IMPLEMENTATION AND UTILISATION OF INTERVENTIONS
1. INTRODUCTION

All of the interventions described in this report are ways in which energy can be used more efficiently to provide a higher level of service at reduced cost. The international energy policy literature has numerous examples of how energy efficiency is often the least-cost path towards providing sustainable energy service (see for example Lovins and Lovins (1991), Reddy and Goldemberg (1990); Gadgil and Jannuzzi (1991) and Kats (1992)). In countries where the gap between access to affordable energy and the demand for clean energy is very large, such as South Africa, energy efficiency has the potential to accomplish multiple social and economic objectives.

Previous South African studies have shown that significant potential for energy efficiency exists across a range of sectors but the costs are not well understood (Thorne 1995). The impacts of energy efficiency on the low-income residential sector are particularly important, because of the social priorities for upliftment and empowerment of the poor. A series of research papers from EDRC have applied traditional CBA to some energy efficiency interventions for the urban poor at a national level (Spalding-Fecher et al (1999); van Horen and Simmonds (1998); Simmonds (1997); Clark (1997); Thorne (1996)). This analysis takes such studies a step further by including a wider range of interventions and a disaggregated analysis at the household level. The basic methodology, however, remains the same.

2. METHODOLOGY FOR COST-BENEFIT ANALYSIS

Figure 3: NPV of interventions at national level and the implications of local externalities
Sources: CBA output of the study

Note that the income groups reported here are for income per person per month. For the affordability analysis, the above data were converted to household income using a factor of six and then further adjusted to derive annual figures used in the analysis. From Table 3.1 it is clear that there is no substantial variation in fuel expenditure between the different income groups except in the highest income category in electrified households. This is supported by several studies such as Simmonds and Mammon (1996) and Afrene-Okese (1998) which show that as household income increases the demand for fuels particularly electricity and fuelwood increases at a proportionally lower rate. Demand for other fuels is reported to be income inelastic.

1. END-USE FUEL EXPENDITURE

In order to assess affordability of an intervention, the expenditure on a particular fuel for a given end-use was required. For example, one needs to know how much a household in Durban might spend per month on electricity for space heating. Although none of the available data sets provides this information in Rands, Simmonds and Mammon (1996, Table 2.11) provides data on total fuel expenditure for different income groups. The data was adjusted to reflect 1999 prices.

Expenditure per fuel on a particular end-use was derived with the assumption that 25% is spent on space heating, 40% on water heating and 5% on lighting (Simmonds & Mammon 1996, Table 5.5). More refined estimates of energy consumption by end-use and climatic region should be addressed through further studies. Multiplying the fuel expenditure per end use and income group (in R/mth) by the savings for a particular intervention, the annual savings on expenditure per intervention on one end-use were derived. Based on these calculations, the payback period and the capital subsidy required were derived.

2. RESULTS OF THE AFFORDABILITY MODEL

The capital subsidy was computed by first establishing the PV of the fuel cost savings at the consumer discount rate over 50 years. The PV was then deducted from the incremental capital cost of the intervention to arrive at the capital subsidy required. Since both the energy savings and the capital costs differ regionally (at least for some interventions), it was necessary to differentiate results for the three regions. The study applies the U1 (Johannesburg) region as an example to illustrate the capital subsidy required as an incentive for households to implement the various interventions.

0. Capital subsidy required

The following two tables present the PV and incremental capital subsidy using the example of the Johannesburg region. Table 27 presents the PV of the intervention for consumers. These values are then deducted from the incremental capital cost (see Table A-1-1) to give the capital subsidy required as presented in Table 28. One should note that the initial incremental capital costs would have to be advanced to households due to lack of household-savings to invest in such interventions. The previous two tables have used the example of one region to illustrate the calculation of PV and capital subsidy in one region. The average capital subsidies that are required at the national level are presented in Table 29 which shows the values averaged over regions.

Table 27: PV per intervention for consumers at 30% discount rate for Johannesburg

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
< R 600/m	355.11	67.81	65.15	698.72	89.48	820.77	209.24	820.77	820.77	123.12	164.15	787.94
< R1 200/m	308.31	58.87	56.57	606.64	77.69	712.60	181.67	712.60	712.60	106.89	142.52	684.10
< R1 800/m	377.76	72.13	69.31	743.28	95.18	873.12	222.59	873.12	873.12	130.97	174.62	838.20
< R2 400/m	384.18	73.36	70.49	755.92	96.80	887.97	226.38	887.97	887.97	133.20	177.59	852.45
< R3 000/m	415.56	79.35	76.25	817.66	104.71	960.49	244.87	960.49	960.49	144.07	192.10	922.07

Source: CBA output of this study

Table 28: Capital subsidy required per household-income category and intervention at 30% discount rate for Johannesburg (Rands)

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
<R600/m	601.89	351.19	296.85	37.28	-682.48	1 060.23	-1 323.24	-925.77	426.23	-20.52	549.85	1 017.06
<R1 200/m	648.69	360.13	305.43	129.36	-670.69	1 168.40	-1 295.67	-817.60	534.40	-4.29	571.48	1 120.90
<R1 800/m	579.24	346.87	292.69	-7.28	-688.18	1 007.88	-1 336.59	-978.12	373.88	-28.37	539.38	966.80
<R2 400/m	572.82	345.64	291.51	-19.92	-689.80	993.03	-1 340.38	-992.97	359.03	-30.60	536.41	952.55
<R3 000/m	541.44	339.65	285.75	-81.66	-697.71	920.51	-1 358.87	-1 065.49	286.51	-41.47	521.90	882.93

Source: CBA output of this study

Table 29: National average capital grant required per household for an income group and per intervention (in Rands)

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
<R600/m	526.91	351.08	288.43	255.01	-664.14	1 060.23	-1 322.90	-925.77	426.23	-20.52	549.85	1 017.06
<R1 200/m	583.59	360.03	298.13	318.40	-654.76	1 168.40	-1 295.37	-817.60	534.40	-4.29	571.48	1 120.90
<R1 800/m	499.48	346.75	283.74	224.33	-668.68	1 007.88	-1 336.23	-978.12	373.88	-28.37	539.38	966.80
<R2 400/m	491.70	345.52	282.41	215.63	-669.96	993.03	-1 340.01	-992.97	359.03	-30.60	536.41	952.55
<R3 000/m	453.69	339.52	275.91	173.13	-676.25	920.51	-1 358.46	-1 065.49	286.51	-41.47	521.90	882.93

Source: CBA output of this study

The variation of capital grants required for different income groups is not great for most interventions. The exception relates to informal houses, where the capital subsidy required to make the package attractive is about twice as high for the poorest households than for those earning between R2 4000 and R3 000 per month. The variation for CFLs is significant in percentage terms, but not in Rands. Capital subsidies are not required for interventions where initial capital is saved. Unsurprisingly, windows, CFLs and the two interventions in the row house, all have a negative subsidy requirement. These are economically attractive discrete interventions, even at a consumer discount rate.

For the 30m² RDP house a capital subsidy of around R1 000 appears to be required to make the package attractive to households. In the context of housing subsidies, this would be a modest amount in view of the substantial economic and environmental benefits. It should be remembered that this is not the full incremental capital cost, but a subsidy which would make the intervention attractive to households. Mechanisms for financing the incremental capital cost (over and above the status quo subsidy) as well as the capital subsidy should be a subject for further studies.

SHS and ceilings have relatively large subsidies due to their relatively high incremental capital cost. One should note that wall insulation requires a modest subsidy, which would differ by region. For example, capital costs in Johannesburg are about double those in Cape Town and more than three times those in Durban.

1. Payback period

The payback period was calculated as the number of periods required to pay back the incremental capital cost of the intervention, using the fuel cost savings as payment. This might be the case where a municipality or utility, for example, pays for the intervention upfront, but continues charging present fuel costs and repays the capital over a period. Here, we assumed that the interest rate was 8%, which would imply that the capital was provided by an institution with a significantly lower discount rate than that of the consumer.

Table 30: Average payback period per income group and intervention (in years)

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
<R600/m	13.12	N/A	N/A	8.36	-15.53	14.97	-12.17	-0.67	6.75	3.27	N/A	12.26
<R1 200/m	17.58	N/A	N/A	10.33	-16.83	25.45	-13.30	-0.77	8.17	3.84	N/A	15.80
<R1 800/m	11.76	N/A	N/A	7.66	-14.98	13.09	-11.69	-0.64	6.23	3.04	N/A	11.10
<R2 400/m	11.43	N/A	N/A	7.49	-14.83	12.66	-11.57	-0.63	6.10	2.99	N/A	10.81
<R3 000/m	10.07	N/A	N/A	6.73	-14.15	10.96	-10.98	-0.58	5.52	2.74	61.43	9.59

Note: ‘N/A’ indicates that the capital amount cannot be repaid in the life of the investment
at that discount rate.

Source: CBA output of this study

Variation between income groups on the payback period is not significant. The effect of the consumer discount rate and its associated time preference appears to be the dominant factor here. The table shows clearly that several interventions have long payback periods (~10 years for the ceiling, wall insulation, the RDP package and the DSWH). In no income group can consumers afford to pay for these interventions, at a discount rate of 30%. For informal houses, the payback period exceeds the estimated life of the shack of five years. Only CFLs show a fairly short payback period.

For some interventions, there is not enough time to make the payments equivalent to energy cost savings in order to repay the capital cost. This is true of roof and wall insulation, as well as SHS. The ‘negative years’ reported in the table have no real-life meaning, but apply to interventions which save on building costs (window and row house). An anomaly should be noted that the lowest income group generally has a lower PV and shorter payback period than the second lowest. This result is probably due to the small sample on which the data in Simmonds and Mammon (1996) is based.

Payback period allows different interventions to be compared in simple terms – how long does it take to pay back the capital cost. This comparison can be made graphically. Fig A-6-1 shows a comparison of payback periods for region U3 (Durban) for the various interventions, for one income group. In the cases where the payback period is negative, this indicates that the capital cost was negative. In other words, there was saving on the initial payment on capital.

Figure 4: Payback periods for interventions in Durban for income group < R600/mth
Source: CBA output of this study

3. CONCLUSIONS AND RECOMMENDATIONS

1. APPLICATION OF AFFORDABILITY ASSESSMENT TO LOW-INCOME HOUSEHOLDS
1. INTRODUCTION

This chapter disaggregates the results from Chapter 2 so as to show how the energy efficiency interventions affect different kinds of households in different regions of the country. Because climate conditions and fuel-use patterns vary considerably, certain interventions could be economic for some groups and not for others. The methodologies used for the cost-benefit analysis are the same as in Chapter 2. The national results in Chapter 2 are the aggregated results from the household level that are presented here.

2. COSTS AND BENEFITS FOR POOR HOUSEHOLDS IN DIFFERENT REGIONS

The value of each intervention per household can be measured in terms of NPV as explained in chapter 3.

1. Calculating NPV per household

The inputs from the housing model, data gathered on fuel use patterns and fuel prices constitutes the basis of this CBA. Fuel consumption figures were collected by region, fuel and end-use, as described in Chapter A-1 These consumption figures were multiplied by the energy savings factor for the various interventions simulated in Chapter A-1 in the wider study to derive energy-saving (kWh, kg or l per year). Energy savings are converted into Rands using respective fuel prices by region. Avoided external costs for the respective intervention are also calculated based on energy savings at the household level.

The benefits to the household are found by taking the present value of both the energy savings and the avoided local external costs (PV 1999Rands). Costs are composed of the incremental capital cost (costs of the intervention, less any avoided costs such as that of an electric geyser), replacement costs and any increased operational costs (such as maintenance requirements) other than fuel costs as described in Chapter A-1).

The NPV for the households adds together the benefits (energy savings and avoided external costs) and deducts the costs (capital, replacement, operational). The present values are calculated at the social discount rate of 8% to provide the social NPV. The consumer NPV is also computed at the consumer discount rate of 30%. Both household level NPVs are disaggregated by 5 fuels and 3 regions.

2. Variation by region/climatic zone

Because of variations in climate and fuel use patterns throughout the country, it is important to examine how the economic benefits vary by region. This value was calculated based on the household NPV for homes using different fuels in each region, weighted by the share of homes using that fuel in each region. Figure 5 and Table 31 below illustrates this variation for each intervention. No external costs are included.



Figure 5: Weighted average regional NPV per household (1999 Rands) including
external costs and aggregating across fuel types
Source: CBA output of this study

Perhaps the most interesting result is how little the NPV varies across regions. This is partly because the regions with the coldest climate, and hence the largest energy savings (Johannesburg), is also in some cases the region with the highest capital costs (e.g. because thicker insulation is required). Part of the variation is also due to the lower prices for electricity in Johannesburg – whose municipalities are closer to the sources of generation and have more industrial customers who can cross-subsidise residential tariffs. This is most evident in the analysis of solar water heaters, where the present value of electricity savings, and hence the NPV varies by as much as R600 across regions. In no cases, however, are there interventions that make sense in one region that do not make sense in another.

3. Effect of external costs

Table 32 and Figure 6 show the impact of external costs on the results. While the interventions clearly have the most economic benefit when we take the external costs of energy into account, the difference is relatively minor, except where the benefit is relatively small (as in DSWH). This is understandable because the majority of the energy savings from these interventions are electricity savings: previous research on external costs of energy has attributed much higher health and environmental impacts to non-electric household fuels than to electricity (Van Horen 1996a; 1996b).

Table 31: Weighted average regional NPV per household (1999 Rands), including external costs and aggregating across fuel types

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
U1 (CT)	950	-329	-240	1 426	687	2 091	1 349	3 362	-675	704	-1 031	818
U2 (Jhb)	708	-308	-256	714	710	893	1 395	2 787	-818	460	-1 174	166
U3 (Dbn)	985	-108	-201	1 043	668	2 018	1 480	3 051	-799	429	-855	265

Source: CBA output of this study

Table 32: Household Net Benefit/Cost with, without and with only local external costs,
for each intervention and aggregating across regions (NPV in 1999 Rands)

	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
Including all externalities	881	-232	-230	1,026	688	1 625	298	3 023	-778	509	-1 010	351
No externalities	672	-266	-266	802	653	1 236	283	2 634	-905	417	-1 123	-22
Local external-ities only	805	-244	-243	938	674	1 480	292	2 878	-826	435	-1 101	51

Source: CBA output of this study

 

Figure 6: Weighted average NPV per household with all, without and with only
local external costs (1999 Rands), aggregating across regions
Source: CBA output of this study

4. Consumer perspective

From a consumer perspective, a higher discount rate should be applied to calculate present values. Where the study addresses the national perspective, a social discount rate (one which reflects both a low risk cost of capital and a social time preference for longer-term investments) is used. This assumes that the future benefits, and even the interests of future generations, are taken into account more strongly than at higher discount rates. If, however, the perspective of the consumer is considered, then the time preference for money (the discount rate) will be higher. In particular, among low-income households where financial resources are extremely limited, the effective discount rate is likely to be much higher – as much as 30% or more (Banks 1999). In plain language, low-income households cannot afford to wait for future benefits and therefore strongly prefer money now to money later. The NPV for households at this discount rate differ substantially from those reported at the social discount rate, above.

Table 33: NPV per household at the consumer discount rate (30%) for each intervention
and region (in 1999 Rands)

Consumer discount rate	Ceiling	Roof ins	Partition	Wall ins	Window	All SH RDP	Shared wall	All SH Row	All SH Inf	CFL	SHS	SWH
U1 (CT)	-481	-395	-333	-212	604	-898	1 146	870	-979	114	-748	-729
U2 (Jhb)	-530	-389	-335	-938	603	-1 716	1 143	669	-1 048	57	-759	-827
U3 (Dbn)	-461	-219	-317	-35	583	-518	1,136	994	-1 022	60	-580	-621

Source: CBA output of this study

Not surprisingly, most of the interventions do not yield a net benefit when a 30% discount rate is assumed – the future energy savings simply have much less value to consumers with high discount rates. In addition, the study has not included external costs in this calculation because these are not attributable to individual households. The reason that changing window size, a shared wall, and the row house still have a positive NPV is that they do not require additional up-front costs – but in fact save money when the house is built. Compact fluorescent lamps, if purchased at the bulk prices that Eskom is projecting for its Efficient Lighting Initiative, are also cost effective even at high discount rate.

3. CONCLUSIONS / RECOMMENDATIONS