Michael Baker - Thesis - Problems in Longterm Forecasting and Planning

About | Contents
Previous: 1. Freight Transport Statistics | This: 2. Vehicle Refuelling Infrastructure | Next: 3. Transport Input-Output Study

2. Vehicle Refuelling Infrastructure

2.1 Introduction

Another avenue of my enquiries started with the Future Transport Fuels (FTF) study (Chapman, Charlesworth and Baker 1976). The FTF study was a short appraisal of the likely alternatives to petrol and diesel as transport fuels in 2025, a notional date by which natural petroleum was assumed to have run out or be in short supply. We looked at many routes between primary sources of energy and transport output. Of these routes the two which seemed most promising in the UK were either liquid fuels derived from coal or electricity stored in batteries [1] . On a primary energy to transport output basis we concluded that the electric vehicle option would be more efficient.

As well as looking at the energy conversion processes directly involved we also examined the choice of transport fuel within the wider energy economy. By examining several energy supply and demand scenarios we concluded that the battery car would be preferable because:

1. In primary energy terms it would be slightly more efficient.
2. The liquid fuelled vehicle scenarios required all the UK's coal for liquid fuel production, whereas in the electric vehicle scenarios there would be sufficient coal available to make gas for heating. Gas is more efficient for heating than the electricity which would be used in the liquid fuel scenario.
3. The seasonal variations in transport demand as compared with space heat demand are such that the electricity system would work at a higher load factor in the electric vehicle scenario.

The FTF study led to further a further study which looked at vehicle refuelling infrastructures (the VRI study). It started from the FTF study and filled in some parts in more detail. The VRI study involved looking at liquid fuel and battery vehicle scenarios (based on those in the FTF study) in some detail to check that there would be no unforeseen problems over their required refuelling infrastructures. It also contained a more detailed analysis of the electricity system operation. The initial aims of the VRI study were to describe and compare in energy and cash terms the refuelling infrastructures required in 2025 by a complete fleet of battery electric road vehicles and of a complete fleet of vehicles running on liquid fuel derived from coal. (The study of the electric vehicle case was to include an examination of improvements required to the grid, however this could not be covered because of the lack of geographical data on road transport.)

Review of other work

In the Electrical Research Association Study reported by Weeks (1978), electric vehicles were to be refuelled only by battery exchange. The reasons given by Weeks for the ERA study being restricted to battery exchange, as the only method of refuelling, were "because home charging would upset the economics of the infrastructure, which would be underutilised, and battery life would be prejudiced by a reduction in quality control of the charging process" (Weeks 1978, p11). However we believed these reasons were wrong because the primary determinant of the economics of exchange stations is their individual turnover rather than how many there are of them, and with suitably designed chargers there should be no problem with the quality of the charging process.

Reasons for conducting the VRI study

The reasons for conducting the VRI study were a desire to check out the implications, in terms of changes to the electricity system as opposed to changes to the liquid fuel manufacturing processes of having one or other type of vehicle fleet. It was recognised that neither a wholly electric nor a wholly liquid fuelled fleet is likely but it was felt that an either/or analysis would highlight the differences between them and that any chosen level of mixture of the two traction types would be arbitrary.

The definition of refuelling infrastructure as used in the VRI study is perhaps best made by reference to Figure 2.1.

Primary Energy						Transport Output
<                                       Refuelling Infrastructure                                             >					<     Vehicle Use    >
coal in ground	extraction	Store - Transport - Store	Conversion No 1	(Store-) Transport (Store)	Store Conversion No 2		trac- tive effort
Stage no	1	2	3	4	5
processes for : electricity from coal, and electric vehicles
	coal mine	storage at mine followed by transport to, and storage at, power stations	conversion of coal to electricity in power station	transmission through grid and battery charging	storage in battery and conversion in electric motor to tractive effort
processes for : liquid fuels from coal
	coal mine	storage at mine followed by transport to, and storage at, coal conversion plant	conversion of coal to liquid fuel in conversion plant	storage of refined product at plant followed by transport to, and storage at, filling station	storage in fuel tank and conversion in internal combustion engine to tractive effort

Figure 2.1 Components of the refuelling infrastructures

This figure illustrates the pathway from primary energy to transport output. This consists of five stages. These are extraction, movement to a conversion plant (power station or 'coalplex'), conversion to a fuel (electricity or liquid), movement to the vehicle, and conversion in the vehicle to tractive effort. Also involved at several points are storage. It is the first four of the stages (up to the point at which fuel is placed in the vehicle) which are considered to be refuelling infrastructure.

In the FTF study we identified significant advantages in other parts of the economy due to electric road vehicles. The allocation of coal as a liquid fuel to road transport could have considerable effect on other sectors which may require other coal derived products such as gas. When coal is in short supply such action (using coal for liquid fuels) may force these other sectors to use a 'second' choice fuel (such as electricity) for some applications ( such as space heating) which leads to a less efficient use of primary energy and higher consumer costs. Consequently when making cost comparisons between the two refuelling infrastructures, comparisons were also made between the total cost of energy within the economy.

2.2 Structure of the VRI study

The VRI study made comparisons between two transport fuel scenarios for the year 2025. On involved an all electric road vehicle fleet and the other an all liquid fuelled fleet. Both scenarios started with vehicle design and performance projections. Official road transport forecasts were used in both scenarios to obtain projected transport energy demands for 2025. Other starting points in the VRI study were projections of energy supply and of energy demands in the remainder of the economy of 2025 plus projected conversion of efficiencies of useful to delivered and of delivered to primary energy. These projections were used in both of the scenarios.

From the starting points the VRI study followed two major strands. The first was the analysis of the two refuelling systems. For electric vehicles it was assumed that the main method of refuelling would be by insitu recharging of batteries and that battery exchange would be used to overcome the range limitations of electric vehicles. To determine the number and size of battery exchange stations required it was necessary to make a detailed analysis of the journey patterns of the different types of road vehicles. The number and size of liquid fuel stations were projected from current trends in size and numbers.

The second strand of the VRI study was an analysis of all energy demands in the UK and of the electricity supply system. First projections of road transport energy demand were made for both electric and liquid fuelled vehicles, and combined with the projected demands in the remainder of the economy. The useful energy demands in the remainder of the economy were broken down into fixed demands, which had to be met from one source, and free demands which could be met from several sources. Fuels were then allocated to road transport followed by each of the fixed demands and then to each of the free demands, up to supply constraints imposed by the projected primary supply availabilities. The demands for electricity were then used in an hourly electricity demand model to determine the required size of generating plant, the breakdown of this plant between coal and nuclear and the amount of coal required for the coal fired plant. It was then necessary to return to the fuel allocation stage to check that the limit on coal production had not been breached.

Finally the two strands were combined to get estimates of fuel costs per vehicle and the total costs of fuels for all energy uses. These cost estimates were made on the basis of notional cost breakdowns (into plant capital and running costs, transmission costs and overheads) for electricity, liquid fuels and gas.

Major assumptions

In developing the two scenarios many assumptions were made. The major assumptions made were:

1. Road transport will follow DoE/DTp forecasts
2. By 2025 there will be no (or negligible) indigenous supply of oil or natural gas.
3. In 2025 the UK will be able to produce 200 to 250 million tonnes of coal per year.
4. The UK will remain energy self sufficient after the depletion of North sea oil and gas.
5. Gross domestic product will grow at the following rates 1975-2000 at 2% and 2000-2025 at 1.5%.
6. There will be a strong preference to use gas for space heating and that the only other fuel available for space heating will be electricity.

   Some of the other assumptions made and developments made upon the original FTF study were:

7. In both scenarios road vehicles would fulfil the same demands and follow the same journey patterns.
8. High energy density batteries will be available with energy densities of 500 MJ/tonne (140 kWh/tonne).
9. Electric batteries will all be hired and not owned.
10. There will be standardisation of vehicles and batteries (in particular size, shape and terminal layout) to allow for battery exchange.
11. Liquid fuels derived from coal will be of the petrol/diesel type which will be stored, dispensed and used in the same way as today.

2.3 Details of the VRI study

Vehicle Design

Unlike the FTF study, the VRI study made and used projections of delivered energy per vehicle kilometre for road transport instead of useful energy per vehicle kilometre. The original intention had been to model vehicle-traffic interactions for bothy electric vehicles and internal combustion engined vehicles. However the modelling of the latter proved difficult and was dropped in favour of making projections of vehicle energy consumptions on the basis of assumed savings over today's levels.

In the VRI study four main categories of vehicle were distinguished (Car, Van, Goods and Bus). For the analysis of the electric vehicles Goods vehicles were further split into Small, Middle and Large corresponding to internal combustion engined goods vehicles of the following Unladen Weights: 1.5-3 tons, 3-5 tons and 5+ tons. The efficiencies of the various parts of the drive system were taken as shown in Table 2.1.

Table 2.1 Electric Vehicle Efficiencies

	efficiency
Battery discharge	0.9
Controller	0.95
Motor	0.92
Transmission	0.95
Overall	0.75

sources:      Corbet and Roerig (1980)
Lee and Corbett (1979)
Charlesworth (1979)

These were combined with the design parameters shown in Table 2.2 and used in a road Vehicle Driving Cycle Model which was principally developed by Charlesworth (1979, pp 176 - 201).

Table 2.2 Electric Vehicle Design Parameters

Vehicle			Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
traction efficiency		%	0.75
battery mass	kg		300	500	900	1400	30000	1400
GVW	kg		1350	2400	6500	11500	35000	13500
payload	kg		400	900	3000	5000	21000	4060
drag coeff.			0.4	0.5	0.85	0.95	1.0	0.85
frontal area	sq.m		1.5	1.5	3.9	7.0	9.0	7.5
Rolling resistance	kgf/kg		0.015	0.013	0.01	0.0075	0.0075	0.0075
battery energy density1	MJ/kg		0.5	0.5	0.5	0.5	0.5	0.5
battery power density1	kW/kg		0.12	0.12	0.12	0.12	0.12	0.12

sources: Charlesworth (1979), Gyenes (1978)

notes: 1 advanced battery such as Sodium-Sulphur or Nickel-Chloride

The model can 'drive' the vehicle over a predefined driving cycle. Those used were the ECE15 Urban, constant speed 56 mph, and constant speed 75 mph cycles (United Nations, Economic Commission for Europe, Inland Transport Committee 1970). The resultant estimates of power requirements, fuel consumption and range for vehicles at full and half load are shown in Table 2.3.

Table 2.3 Estimates of Power, Fuel Consumption and Range

	ECE Driving Cycles, Full load
Vehicle				Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Maximum Power kW
required				17	22	84	162	275	158
available				36	60	108	168	360	168
Fuel Consumption MJ/km (in vehicle energy)
Urban				0.42	0.66	1.70	2.76	7.43	3.12
Constant 56mph				0.43	0.58	2.07	3.86	6.17	3.82
Constant 75mph				0.67	0.87	3.37	6.47	9.72	6.33
Range		km
Urban				360	378	265	253	202	224
Constant 56mph				349	434	218	181	243	183
Constant 75mph				225	287	133	108	154	111
	ECE Driving Cycles, Half load
Maximum Power kW
required				16	21	82	159	230	156
available				36	60	108	168	360	168
Fuel Consumption MJ/km (in vehicle energy)
Urban				0.36	0.55	1.37	2.29	5.44	2.73
Constant 56mph				0.41	0.54	1.97	3.73	5.66	3.72
Constant 75mph				0.65	0.83	3.27	6.35	9.20	6.22
Range		km
Urban				414	454	328	306	276	256
Constant 56mph				366	465	229	187	265	188
Constant 75mph				232	300	137	110	163	112

The three sets of fuel consumption and range estimates were averaged as shown in  which also includes allowances for accessories and heating. This averaging was for vehicles with half load and took account of the proportion of travel under the three conditions.

Table 2.4 Electric Vehicle Energy Consumption and Range

Vehicle1					Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Proportion of traffic2
Urban					0.496	0.494		0.323		0.568
Rural	(constant 56mph)				0.418	0.441		0.677		0.363
Motorway	(constant 75mph)				0.086	0.065		-3		0.069
Fuel Consumption				MJ/km	0.41	0.56	1.78	3.26	5.59	3.33
Accessories				MJ/km	0.04	0.04	0.04	0.04	0.04	0.06
Total4				MJ/km	0.45	0.60	1.82	3.30	5.63	3.39
Heating5				MJ/km	0.09	0.08	0.05	0.05	0.10	0.15
Energy Stored				MJ	150	250	450	700	1500	700
Average Range				km	340	410	250	210	266	210

notes:     1 all vehicles at half load
2 from Department of Transport (annual a)
3 goods vehicles taken at 56 mph on motorways
4 in vehicle energy
5 average in vehicle energy, supplied by a liquid fuel

The maximum demand for accessories is likely to be 550W (being 250W for Headlights, sidelights, flashers, panel lights etc., and 300W for wiper motor, heater blower etc.). In view of the short time that these accessories are in use it was assumed that they would add an additional energy demand of 0.03-0.06 MJ/km. Because the internal combustion engine acts as a free heat source there has to date been on incentive to reduce heat losses in road vehicles. Instead large heaters are used (typically 6-9kW which compares with the average domestic central heating system which is also rated at 6-9kW). The electric vehicle has no such 'free' heat source and will certainly be better insulated and use heat recovery on ventilated air. Estimates of average heating requirements are given in Table 2.4. These compare with the requirements for currently available heaters (Yonwin 1979), and were assumed to be provided for by a liquid or gas derived from coal.

In the VRI study liquid fuelled vehicles were broken down into groups. (Cars by engine size, Vans and Goods vehicles by unladen weight and Busses by seating capacity). Present fuel consumptions, assumed improvements in efficiency and the resultant future fuel consumptions are shown in Table 2.5. The improvements in efficiency are likely to be due to the use of lighter materials to reduce body mass, better styling to reduce air drag and better tyres and braking systems to reduce rolling resistance. Also shown in Table 2.5 are normalized population breakdowns for 1974 and projected for 2025. It was assumed that, except for goods vehicles, within each group of vehicles the average range is the same. (For example see the weekly mileages of cars by engine capacity in Gray (1969)). Consequently the normalised populations also represent normalised total annual ranges and can be used as weighting factors to find average fuel consumptions.

Table 2.5 Present and Projected Liquid Fuelled Vehicle Consumption

		Normalised population		Fuel Consumption
Vehicle	group	1976	20251	Present MJ/km	Improve- ment %2	Future MJ/km
Car  by  engine  size  litre	<1	0.16	0.15	2.6	40	1.6
	1-1.5	0.46	0.50	3.4	40	2.0
	1.5-2	0.30	0.32	4.1	40	2.5
	2-3	0.05	0.02	5.3	40	3.2
	3+	0.03	0.01	6.3	40	3.8
	Total/av	1.00	1.00	3.66	41.5	2.14
Van  by  ULW  ton	<0.8	0.38	0.25	3.0	40	1.8
	0.8-1	0.16	0.24	3.2	40	1.9
	1-1.5	0.46	0.51	4.4	40	2.6
	Total/av	1.00	1.00	3.68	38.6	2.26
Goods  by  ULW  ton	1.5-2	0.11	0.10	5.1	20	4.1
	2-3	0.23	0.23	7.3	20	5.8
	3-5	0.26	0.10	8.7	20	6.7
	5-8	0.22	0.32	12.5	20	10.0
	8+	0.18	0.20	16.0	20	12.8
	Total/av	1.00	1.00	3	3	3
Bus  by  seats	8-32	0.12	0.12	7.3	20	5.8
	32-48	0.30	0.28	10.6	20	8.5
	48+	0.58	0.60	13.2	20	10.6
	Total/av	1.00	1.00	11.71	19.6	9.41

notes:         1 projected
2 after Waters and Laker (1980) and Waters (1980)
3 range varies with size (see Table 2.7)

Transport Projections

Projections of vehicle numbers and total kilometres were made by extending the DoE/DTp forecasts (Department of Transport annual a) from 2010 to 2025. The lower forecasts were used because these allow the best correspondence between GDP assumptions used in the VRI study and the forecasts (for example see Tanner (1974, p6). The results of doing this are shown in Table 2.6.

Table 2.6 Vehicle and Vehicle Kilometre Projections for 2025

Vehicle	Number of vehicles million	Total vehicles kilometres '000 million
Car	26.4	418
Van	2.0	51
Goods	1.1	49
Bus	0.09	3.4

source: after Department of the Environment (1975c)

Energy Demand Projections

Average goods vehicle fuel consumptions for both electric and liquid fuelled vehicles were found by using 1974 average annual ranges (Department of Transport 1979a) in conjunction with the projected 2025 population breakdown to derive normalized average annual ranges for 2025. These together with projected numbers and vehicle kilometres for goods vehicles are shown in Table 2.7.

Table 2.7 Goods Vehicles, Vehicle km and Fuel Consumptions

Size				Small		Medium	Large		Total/av
ULW	ton			1.5-2	2-3	3-5	5-8	8+	1.5+
Vehicle km/yr (1974)		'000		19	19	27	30	50	-
normalised population (2025)				0.10	0.23	0.10	0.32	0.25	1.00
Number of vehicles	million			0.11	0.25	0.11	0.35	0.28	1.1
normalized vehicle km (2025)				0.06	0.14	0.09	0.31	0.40	1.00
total vehicle km '000	million			3.0	6.8	4.2	15.0	20.0	49
vehicle km/yr (2025)		'000		27.2	27.2	38.7	43.0	71.7	-
Liquid Fuel Consumption			MJ/km	4.1	5.8	6.7	10.0	12.8	9.88
Electrical Energy	MJ/km			1.82		3.30	5.63		4.66
Heating	MJ/km			0.05		0.05	0.10		0.09

Total projected fuel consumption for both electric and liquid fuelled vehicles are shown in Table 2.8.

Table 2.8 Projected Vehicle Fuel Consumptions

	Total Vehicle km/yr '000 million	Liquid Fuelled		Electric Vehicles
Vehicle		consumption		electric energy		heating
		MK/km	PJ/yr	MJ/km	PJ/yr	MJ/km	PJ/yr
Car	418	2.14	895	0.45	188	0.09	38
Van	51	2.26	115	0.60	31	0.08	4
Goods	49	9.88	484	4.66	228	0.09	4
Bus	3.4	9.41	32	3.39	12	0.15	1
Total			1526		495		47

These were found by combining the previously found fuel consumptions with total vehicle kilometres.

Analysis of Journeys

Because of the limited ranges of the electric vehicles they would be unable to do all today's journeys on one battery charge. Consequently it was assumed that, to enable electric vehicles to perform journeys of any length, battery exchange stations would be provided. To find the number of exchanges required on any day, it is necessary to know the distribution of distances covered on that day, and average distance travelled before an exchange is required. For this purpose 'long journey' range was postulated which is somewhat shorter than the average range due to assumed higher average speeds on long journeys. The long journey range was taken to be the average of the 56mph and 75mph ranges for cars, vans and buses and the 56mph range for goods vehicles. The average 'exchange range' was then taken to be this 'long journey' range less 20km (0.5 of an assumed 20km exchange station spacing, plus a 10km safety margin.)  (Later in the study it was found that 2500 stations would be required. These could be spaced on average at 19km on all trunk and principal roads (Department of Transport annual a). These ranges plus the average discharge factor (a measure of how far the average battery is discharged on exchange) are shown in Table 2.9.

Table 2.9 Exchange Ranges for Electric Vehicles

kilometres
vehicle1	Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Urban	414	454	328	306	276	256
Constant 56mph	366	465	229	187	265	188
Constant 75mph	232	300	137	110	163	112
"Long journey"2	300	380	230	190	260	150
"exchange"	280	360	210	170	240	130
av discharge factor	0.93	0.95	0.91	0.89	0.92	0.87

notes:         1 all vehicles at half load
2 car, van and bus at average of 56mph & 75mph, goods at 56mph

To obtain an estimate of the total number of battery exchanges per year the number required on an average day was estimated. To estimate the size of stations required, the number of exchanges required on a peak day were estimated. The peak is likely to occur either on a day of maximum car use or a day of maximum goods vehicle use. Historically these have been a Saturday in August and an October weekday respectively.

The distribution of daily ranges were estimated by taking historic distributions of ranges and adjusting them so that they coincide with the projected number of vehicles in use and the average daily kilometres for each type of vehicle for each of the three days. The proportion of vehicles in use and average daily mileages are shown in Table 2.10.

Table 2.10 Summary of Vehicle Use and Average Daily Mileages

vehicle			Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
number	million (1)		26.4	2.0	0.36	0.11	0.63	0.09
Total annual km	'000 million (2)		418	51	9.8	4.2	35.0	3.4
Average daily km		(3)	43.4	70	75	105	152	100
Ratio of day's   to av km	Aug. Sat (4)		1.42	0.82		0.34		1.34
	Oct. weekday (5)		0.98	1.17		1.34		1.13
Proportion   of vehicles   in use	Aug. Sat (6)		0.8	0.53		0.21		0.8
	Oct. weekday (7)		0.75	0.75		0.8		0.8
	Av. Day (8)		0.74	0.63		0.59		0.75
Av daily km   of vehicles   in use	Aug. Sat (9)		77	110	127	178	258	170
	Oct. weekday (10)		57	110	127	178	258	140
	Av. Day (11)		59	110	127	178	258	130

notes:    row (1) from Table 2.6 }
row (2) from Table 2.6 } Goods from Table 2.7
row (3) = row (2) / row (1) / 365000
rows (4) & (5) derived from "Monthly and average daily traffic on classified roads: by type of vehicle: 1974" (Department of Transport annual a)
rows (6) to (8) derived from Gray (1969), Department of the Environment (1975b) and Edwards and Bayliss (1970)
The figures for Van And Goods are such that the average daily km for the 3 days in rows (9) to (11) are the same
row (9) = row (3) * row (4) / row (6)
row (10) = row (3) * row (5) / row (7)
row (11) = row (3) / row (8)

The distribution of daily ranges for cars was derived from data in Gray (1969) and the 1972-73 National Travel Survey (Department of the Environment 1975a). The distribution of annual and weekly mileages when reduced to a common basis were found to be very similar and it was assumed that the distribution of daily mileages has the same shape. Less data is available for vans and goods vehicles. It was assumed that they will have distributions similar to the distributions of journey lengths. These were obtained from the Sample of Small Goods Vehicles: 1974 (Department of the Environment 1975b) and the Survey of Road Goods Transport 1962 (Ministry of Transport 1964-66). The distribution of daily ranges of buses was obtained from Hellewell (1978).

The derived distributions of daily mileages are shown in Figure 2.2 to Figure 2.7. On these figures are shown the exchange ranges of each type of vehicle. The proportion of vehicles requiring 1st, 2nd, etc. exchanges, on each of the three days considered, are shown in . The final column in this table gives an estimate of the number of exchanges required as a proportion of each type of vehicle. It should be noted that these estimates are liable to error as they are based upon the size of the tails of very imperfectly known distributions.

Figure 2.2 Cumulative distributions of daily ranges for cars

Figure 2.3 Cumulative distributions of daily ranges for vans

Figure 2.4 Cumulative distributions of daily ranges for small goods

Figure 2.5 Cumulative distributions of daily ranges for middle goods

Figure 2.6 Cumulative distributions of daily ranges for large goods

Figure 2.7 Cumulative distributions of daily ranges for busses

Table 2.11 Proportion of Vehicles Requiring Exchanges

August Saturday
vehicle	Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
1st exchange	0.01	0.04	0.036	0.06	0.07	0.32
2nd exchange	-	0.01	0.014	0.03	0.04	0.13
3rd exchange	-	-	0.006	0.02	0.03	-
4th exchange	-	-	0.004	0.02	0.02	-
5th exchange	-	-	-	0.01	0.01	-
6th exchange	-	-	-	0.01	-	-
exchanges required as proportion of vehicles	0.01	0.05	0.06	0.15	0.17	0.45
October Weekday
vehicle	Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
1st exchange	0.005	0.06	0.14	0.23	0.25	0.27
2nd exchange	-	0.02	0.055	0.12	0.15	0.05
3rd exchange	-	-	0.025	0.08	0.10	-
4th exchange	-	-	0.0	0.05	0.05	-
5th exchange	-	-	-	0.04	0.02	-
6th exchange	-	-	-	0.03	-	-
exchanges required as proportion of vehicles	0.005	0.08	0.23	0.55	0.57	0.33
Average day
vehicle	Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
1st exchange	0.005	0.05	0.10	0.17	0.19	0.24
2nd exchange	-	0.02	0.04	0.09	0.11	0.03
3rd exchange	-	-	0.02	0.06	0.07	-
4th exchange	-	-	0.01	0.04	0.04	-
5th exchange	-	-	-	0.03	0.01	-
6th exchange	-	-	-	0.02	-	-
exchanges required as proportion of vehicles	0.005	0.07	0.07	0.41	0.42	0.27

Insitu Recharge Requirements

Currently available battery chargers generally need a transformer to either step the voltage down or up. For a 1 to 20 kW charger they cost of the order of £₁₉₇₅ 50 to 100 per kW. They are of such a size and weight that they could not be carried in vehicles. For the purposes of the analysis, and to demonstrate that a transformerless on board charger would be relatively cheap and simple, a simple recharge system was envisaged. The advantages of an on board charger are that recharging could be done anywhere there was a suitable supply. For example by cars at home or possibly in car parks or at parking meters. This would place no special requirements upon domestic supplies. The current required would be 30 to 40 amps and electric cookers age generally connected up by 60 amp cable. However difficulties would arise over a plug and socket able to carry this much current. It would have to involve compression of the conductors but given the potential mass market (approx. 20 million) should be reasonably cheap to produce.

The simple design of charger would have batteries with full charged voltages just below mains voltage (240V). Charging would start with a peak power requirement and carry on until the battery reaches this voltage at which point there would be a very small power requirement. From a constant voltage source the charge entering a battery per second is an exponentially decaying function since the rate of flow depends upon the difference in voltage between source and battery. Due to the finite internal resistances of the source, battery and charger, the rate of charging approximates to a straight line function. If

I = peak current

V_s= supply voltage

t_c = charge time

E = energy stored in a fully charged battery

P = peak power raring of charger (= IV_s)

Then E = 1/2 I V_s t_c = 1/2 P t_c

∴ P = 2 E / t_c

To charge a 40 kWh battery (E) in 8 hours (tc) would require a charger with a 10kW raring (P). The charger power ratings for the batteries of each of the vehicles considered at various recharge times are shown in Table 2.12.

Table 2.12 Insitu Recharge Power Requirements

vehicle				Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
battery size			kWh	42	69	125	194	417	194
Recharge time of		4hr	kW	21	35	63	97	209	97
		8hr	kW	11	17	31	49	104	49
		12hr	kW	7	12	21	32	70	32

Further analysis of the systems effects of recharging are covered by the electricity system analysis (see below).

Battery Exchange Stations

In this part  of the analysis the size of battery exchange infrastructure required to meet the previously found demand for battery exchanges and the costs of exchanges (net of taxes and profits) were estimated. The proportion of vehicles requiring exchanges (Table 2.11) were converted to numbers of exchanges required and the amount of energy dispensed on the three days was found (see Table 2.13). The peak demand for batteries and for energy is on an October weekday due to goods vehicles.

Table 2.13 Battery Exchange Requirements

vehicle			Car	Van	Small Goods	Mid. Goods	Large Goods	Bus	Total
No of Vehicles		million	26.4	2.0	0.36	0.11	0.63	0.09
No of Exchanges '000/day1	Aug. Sat		264	100	22	17	107	41	551
	Oct. weekday		132	160	83	61	359	30	825
	Av. Day		132	140	61	45	265	24	667
Energy Delivered TJ/day2	Aug. Sat		37	24	9	11	148	25	254
	Oct. weekday		18	38	33	38	495	18	640
	Av. Day		18	33	25	28	366	15	485

notes:   1 No of vehicles * exchanges required as proportion of vehicles (see Table 2.11)
2 Battery size (see Table 2.12) * number of exchanges * average discharge factor (see Table 2.9)

A hypothetical exchange station was envisaged with two car/van bays and two goods/bus bays with fully automated exchange mechanisms for cars and semi-automated exchanges for goods. On the assumption that each exchange would take 5 minutes, the station could complete 24 car/van exchanges and 24 goods/bus exchanges per hour. With a 12 hour day and 360 thousand car/van exchanges on the August Saturday would require 1250 stations and the 530 thousand goods/bus exchanges on an October weekday would require 1440 stations on the assumption that there were no private exchange facilities. To allow for breakdowns and regional variations it was assumed that 2500 stations would be required.

The numbers of exchanges made on an average and a peak day by a typical station are shown in Table 2.14. With an 8 hour recharge period and the maximum energy dispensed of 71 MWh per station (640 TJ/2500 stations, see ) might require a peak charger capacity of 18MW. However this would only occur if all batteries were to start charging at the same time. It seems likely that the start of charge would be staggered over an average of perhaps four hours which would reduce the peak charger capacity required to 9MW. For the hypothetical station a 10MW charger was chosen. This would probably require its own substation.

Table 2.14 Battery Stock Requirements

vehicle	Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Average day	53	56	24	18	106	10
peak day1,2	127	77	40	29	172	20

notes: 1 The peak day occurs at different times for different vehicle types
2 Including a 20% safety margin

The costs of running an exchange station were broken down into four components: operating costs, station capital charges, electricity costs, and battery capital charges. The first two were distributed over all users on an equal basis per exchange and the last two were attributed to each vehicle individually. Operating costs were taken to be £₁₉₇₅ 15,000 for staff (forecourt attendant, cashier, manager and three mechanics) plus £₁₉₇₅ 29,000 for rates, station maintenance etc. Station capital cost estimates are shown in Table 2.15, and are comparable to those given by Weeks (1978). The capital costs of batteries would be quite high with car batteries of the order of £₁₉₇₅ 840 and large goods at £₁₉₇₅ 8,300 (approx. £20/kWh, see Weeks 1978, p8). To avoid the necessity for transfer payments on each battery exchange the batteries would have to be hired. The annual hire charges for various interest rates are shown in Table 2.16. The total annual capital charges for the station battery stocks are shown in Table 2.17. Finally the fuel costs of recharging are shown in Table 2.18. Because battery recharging can be used to fill the troughs on the electricity system (see electricity system analysis below) it was assumed that there will be a discount from the average cost of 2.1 p/kWh () to 1.4 p/kWh. The total of all four cost elements for each vehicle type are shown in Table 2.19.

Table 2.15 Exchange Station Capital Costs

£'0001975
Site	(650 sq m @  £20/sq m)	13
Buildings	(125 sq m @ £175/sq m)	22
Equipment:
chargers and electricity   supply (10MW @ $10/kW)		100
exchange equipment		150
workshops		100
Total		385

Table 2.16 Annual Battery Hire Charges

vehicle		Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Annual kilometres		15800	25500	27200	38700	55555	37777
Battery life1	years	20	12.5	11.8	8.3	5.8	8.5
Size of Battery	kWh	42	69	125	194	417	194
Capital Cost2	£1975	840	1380	2500	3880	8340	3880
Annual hire charge3 £1975/yr
Interest Rate %	5	67	151	286	583	1692	571
	7	79	169	318	632	1799	621
	10	99	198	370	710	1964	699
	15	134	251	464	848	2252	837
	20	172	307	566	995	2556	985

notes:  1 @ 320 000 km (Anon 1980)
2 @ £20/kWh (Weeks 1978)
3 repayment = Capital * r(1+r)ⁿ/(1+r)^n-1

Table 2.17 Battery Stock Capital Charges per Station

vehicle		Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
No of batteries		127	77	40	29	172	20
Annual Capital Charges £'0001975	5%	8.5	11.6	11.4	16.9	291.0	11.4
	10%	12.6	15.2	14.8	20.6	337.8	14.0
	20%	21.8	23.6	22.6	28.9	429.6	19.7

Table 2.18 Battery Recharge Average Fuel Costs

vehicle		Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Battery size	kWh	42	69	125	194	417	194
Av discharge factor1		0.93	0.95	0.91	0.89	0.92	0.87
Cost per exchange2	£1975	0.55	0.92	1.59	2.42	5.37	2.36

notes:  1 From Table 2.9
2 @ 1.4 p/kWh

Table 2.19 Estimated Average Cost of Exchanges (net of profits and taxes)

vehicle			Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Station Capital Charges1	(10%-10yrs)		0.64
Operating costs1			0.12
Battery Capital Charges		(10%)	0.65	0.74	1.69	3.14	8.73	3.84
Fuel costs2			0.55	0.92	1.59	2.42	5.37	2.36
Total Cost £/exchange			1.96	2.42	4.04	6.32	14.86	6.96

notes:  1 equal cost per exchange (Annual cost/ av exchanges per day (=267)/ 365)
2 From Table 2.18

Liquid Fuelled Stations

As with battery exchange stations estimates were made of the required number of filling stations and the cost of dispensing fuel (net of profits and taxes). In this case the analysis was easier because it was assumed that the liquid fuels derived from coal could be handled, stored and dispensed in the same way as petrol and diesel are today. Having examined recent trends in the average size of filling stations (210,000 gal/yr in 1975, see Figure 2.8) it was estimated that 500,000 gal/yr stations would be typical for 2025. A further simplifying assumption was made that all fuel would be dispensed at filling stations to obviate the need to estimate quantities which will be privately handled at Goods and Bus owners own depots. For the 1526 PJ/yr (Table 2.8), which is 9.6 * 109 gallons per year, there would need to be approximately 23,000 stations.

Figure 2.8 UK Petrol sales and numbers of filling stations

Estimates of filling station capital costs are given in Table 2.20 for a typical 500,000 gal/yr station. These estimates were based upon Weeks (1978). The annual station running costs are given in Table 2.21. The total cost of £₁₉₇₅ 38,000 per year is equivalent to 8p/gallon.

Table 2.20 Filling Station Capital Cost

(for 500,000 gal/yr self service station)

£'0001975
Land	14
Buildings and Services	88
Tanks and Pumps	12
Total	114

source: after Weeks (1978)

Table 2.21 Average Filling Station Running Cost

(for 500,000 gal/yr self service station)

£'0001975/yr
Capital Charges (10%-10yrs)		19
Labour		13
Rates, Insurance, and other costs		8.6
Total		38
Cost	(p/gallon)	7.6
	(£/GJ)	0.48

source: after Weeks (1978)

UK Primary energy supply

Following an analysis similar to that in the FTF study, the primary energy availability shown in Table 2.22 was estimated. It was assumed that following the decline of north sea oil and gas the UK will remain energy self sufficient so for 2025 these would be the only supplies available.

Table 2.22 Projected UK Primary Energy Supply for 2025

	mtce/yr
Coal	200-250
Oil	0
Gas	0
Primary Electricity
Nuclear	211-422
Waves	50
Solar	40

based on:    Department of Energy (1976),
National Coal Board (1974),
National Coal Board (1974)

Useful Energy Demand Projections

The two road transport fuel scenarios were compared in terms of national energy demand. This was found by making useful energy demand projections and then meeting these demands by allocating fuels up to limits imposed by primary energy availability.

The projections of useful energy were taken from the FTF study. They were for the six sectors: Domestic, Public Services, Other Services, Agriculture, Industry and Transport (other than road). Road transport was dealt with separately as noted above. The projections were made by relating the historic demand for useful energy to activity levels for each sector. For domestic the activity level used was houses and for the remainder the contribution each sector makes to GDP (net output). The activity levels were then projected to 2025 to get useful energy demands. The growth rates for GDP assumed were for 1975 to 2000 2% and for 2000 to 2025 1.5%. These projections of the historic relationship between activity level and useful energy take no account of likely reductions due to conservation so percentage reductions were made as shown in Table 2.23.

Table 2.23 Useful Energy Demand Projections for 2025

Sector					Domestic	Public Services	Other Services	Agriculture	Industry	Trans- port1
Activity Indicator					Number of dwellings	G.D.P. Contribution (net output)
Gradient
	Useful energy per unit activity GJ/dwelling or MJ/£1970				58	30	10	30	80	45
Intercept at zero activity			PJ		0	50	80	-10	0	20
Number of dwellings or Proportion of G.D.P.2					30	0.23	0.38	0.03	0.33	0.03
Useful energy demand		without Conservation		PJ	1740	930	560	100	3410	220
		Conservation saving		%	25	25	25	-	28	15
		with Conservation		PJ	1300	700	420	100	2500	190

source:   Chapman, Charlesworth and Baker (1976)

notes:   1 Except for Road Transport
2 GDP growth rates assumed 1975-2000 2.5% pa and 2000-2025 1.5% pa, giving GDP in 2025 of £'000 million 127.8

Fuel Allocations

Using the same procedure as used in the FTF study, fuels were allocated first to Road Transport then to fixed demands and finally to flexible demands. The breakdown of useful energy to fixed and flexible demands is shown in Table 2.24.

Table 2.24 Fixed and Flexible Useful Energy Demands for 2025

PJ
Sector		Domestic	Public Services	Other Services	Agricul- ture	Ind- ustry	Trans- port1
Fixed   demand   for	Coal	-	-	-	-	1000	-
	Gas	100	-	-	-	-	-
	Liquid	-	-	-	60	-	130
	Solar	280	150	100	-	-	-
	Electricity	330	210	126	-	750	60
Flexible demand		590	340	194	40	750	-
Total useful   energy demand		1300	700	420	100	2500	190

source: Chapman, Charlesworth and Baker (1976)

notes:  1 Except for Road Transport
 

The conversion efficiencies of Primary energy to fuel and fuel to useful energy are shown in Table 2.25 and Table 2.26.

Table 2.25 Primary to Delivered Energy Conversion Efficiencies

	Coal	Coal to				Nuclear waves1		Solar
		Liquid	Gas	Elect- ricity	Bat- teries	Elect- ricity	Bat- teries	Heat2
Extraction	0.95	0.95	0.95	0.95	0.95
Conversion	-	0.65	0.70	0.35	0.35
Transmission	-	0.903	0.98	0.90	0.90
Battery  charging	-	-	-	-	0.80
Overall  efficiency	0.95	0.56	0.65	0.30	0.24	0.30	0.24	0.65
Fuel per unit of primary energy PJ/mtce	14.66
	27.36	liquid 1.47	18.72	8.64	6.91	8.64	6.91	18.72
		gas

source: Chapman, Charlesworth and Baker (1976)

notes:   1 on a coal equivalent basis
2 on a gas equivalent basis
3 refining and delivery
 

Table 2.26 Delivered to Useful Energy Conversion Efficiencies

Fuel	Coal	Liquid	Gas	Electricity
Heat	-	-	0.7	1.0
Light, Work	-	-	-	1.0
Transport	-	0.2	-	-
Agriculture	-	0.3	-	-
Industrial  Carbon	0.7	-	-	-

source: Chapman, Charlesworth and Baker (1976)
 

The order of allocation was:

1. allocate fuel(s) to road transport (see Table 2.27)
2. allocate fuels to fill the fixed demands: industrial carbon, gas to the domestic sector, liquid fuels for agriculture and other transport and electricity in all sectors except agriculture.
3. allocate by-product gas (from liquid fuel production) then gas from coal (up to a nominal coal limit of 250 million tonnes) to fill flexible demands and finally to fill any remaining demands with electricity.

Table 2.27 Primary Energy Demand for Road Transport

Vehicles	Liquid fuelled		Electric
			electricity		heat
	delivered PJ/yr	primary mtce/yr	delivered PJ/yr	primary mtce/yr	delivered PJ/yr	primary mtce/yr
Cars	895	61	188	27	38	2.5
Vans	115	8	31	4	4	0.3
Goods	484	33	228	33	4	0.3
Busses	32	2	12	2	1	0.1
Total	1526	104	459	66	47	3.2

In the electric vehicle scenario (Table 2.28) it was possible to meet all flexible demands from coal. However for the liquid fuelled vehicle scenario (Table 2.29) there were only 16 million tonnes of coal remaining after fixed demands had been met. This was reserved for electricity production (see below), and the flexible demands were met from electricity.

Table 2.28 Primary Energy Allocation: Electric Vehicle Scenario

mtce/yr
	Coal
Sector	Coal	Liquid	Gas	Solar	Electricity	Total
Domestic	-	-	[6]147	21	38	106
Public Services	-	-	26	11	24	61
Other Services	-	-	15	8	15	38
Agriculture	-	14	-	-	5	19
Industry	52	-	57	-	87	196
Road Transport	-	3	-	-	66	69
Other Transport	-	44	-	-	7	51
Coal for electricity	0	-	-	-	0	0
Total	258			40	242	540

notes:     1 by-product gas which is not counted in total demand

Table 2.29 Primary Energy Allocation: Liquid Fuelled Vehicle Scenario

mtce/yr
	Coal
Sector	Coal	Liquid	Gas	Solar	Electricity	Total
Domestic	-	-	[16]1	21	94	115
Public Services	-	-	-	11	64	75
Other Services	-	-	-	8	37	45
Agriculture	-	14	-	-	5	19
Industry	52	-	-	-	174	226
Road Transport	-	104	-	-	-	104
Other Transport	-	44	-	-	7	51
Coal for electricity	44	-	-	-	-44	0
Total	258			40	337	540

notes:     1 by-product gas which is not counted in total demand

Electricity System

A half hourly electricity demand model (see Appendix 4) was used to examine the effects of load control by the grid system operators and of a system store on the average load factor and the installed capacity. By load control is meant any method of interrupting a load so that it fills in the trough in each day's demand. This can be achieved in any of several ways such as ripple control or logic devices in each appliance (Salter 1977). On the assumption that the latter would be used, any load to which load control is applied is referred to as a 'logic load'. The electricity model was also used to determine the nuclear and coal fired capacities required.

The model used as input the previously found electricity demands from each sector. These are shown in Table 2.30. The model has three main categories of demand (Domestic, Commercial and Industrial) to which was added battery recharging and off peak heating. These two additional loads are potentially 'off-peak' loads and can be given a daily load pattern similar to today's off-peak domestic space heating load or can be treated as a logic load. For the liquid fuelled vehicle scenario loads larger than those in the electric vehicle scenario were taken to be space heating and  suitable for treatment as potential 'off-peak' loads (see Table 2.30).

Table 2.30 Electricity Demands

Scenario	Electric vehicles				Liquid fuelled vehicles
	mtce		GWyr1		mtce		GWyr1
Sector			"on-peak"	"off-peak"			"on-peak"	"off-peak"
Domestic		38	12	-		94	12	17
Services:
Public	24}	51	15	-	64}	113	15	18
Other	15}				37}
Agriculture	5}				5}
Other Transport	7}				7}
Industry		87	26	-		174	26	26
Road Transport		66	-	20		-	-	-
Total "off-peak"				20				61

notes:   1 With 95% mining & 35% generating efficiency 1 mtce will produce 28.8*10¹⁵*0.95*0.35 / (3.6*10¹²*8760) = 0.304 GWyr of electricity at the power station.

For the electric vehicle scenario the results of running the model with and without a system store and with and without battery charging on logic load are illustrated in Figure 2.9 and Table 2.31.

Table 2.31 Electricity System for Electric Vehicle Scenario

	Without logic load		With logic load
	Without System Store	With System Store	Without System Store	With System Store
Size of System Store            GWh	-	269	-	15
Maximum output of store          GW	-	53	-	5
Average load on System           GW	73	75	73	73
Maximum load on System           GW	123	80	83	79
Load factor       %	59	94	88	93

With battery charging on the 'off-peak' shape of load, charging would cause the system peak of 123GW, so this case is inconsistent. With charging on logic load the maximum load on the system would be 83GW which could be reduced to 97[sic]GW with a system store of 15GWh.

Figure 2.9 Load duration and load patterns for electric vehicle scenario

For the liquid fuelled vehicle scenario the results of running the model are illustrated in Figure 2.10 and Table 2.32. As with the electric vehicles an 'off-peak' shape to the heating load would produce a very large peak of over 500GW. Consequently such 'off-peak' heating will not exist. Of the four cases the lowest maximum load is for the logic load heating at 199GW. The introduction of a system store would have no effect on the maximum load and would not be worthwhile.

Figure 2.10 Load duration and load patterns for liquid fuelled vehicle scenario

Table 2.32 Electricity System for Liquid Fuelled Vehicle Scenario

		Without logic load		With logic load
		Without System Store	With System Store	Without System Store	With System Store
Size of System Store	GWh	-	1765	-	84
Maximum output of store	GW	-	336	-	10
Average load on System	GW	114	122	114	115
Maximum load on System	GW	551	215	199	199
Load factor	%	21	57	57	58

The optimal supply systems for the two scenarios are assumed to be nuclear plants only with system storage for the electric vehicle scenario and a mixture of coal fired and nuclear plant with no system storage for the liquid fuelled vehicle scenario. As previously noted there is a limited supply of coal for the second case and it would all be used by 70GW of coal plant. The installed capacities required for the two scenarios are shown in Table 2.33.

Table 2.33 Electricity System Installed Capacity

Scenario		Electric Vehicles	Liquid Fuelled Vehicles
Maximum load	GW	79	199
Planning margin (20%)	GW	16	40
Installed capacity	GW	95	239
Store Size	GW	5	-
Coal fired plant	GW	-	70
Nuclear plant	GW	95	169
Electricity generated
Coal	GWyr/yr	-	131
Nuclear	GWyr/yr	73	101
Average plant load factors
Coal	%	-	19
Nuclear	%	77	60

notes:   1 44 mtce/yr will provide 44 * 0.304 = 13 GWyr/yr (see Table 2.30, note 1)
 

Fuel Costs

Following the fuel cost estimates made in the FTF study, estimates were made of electricity, gas and liquid fuel costs as shown in Table 2.34. For nucelar generaged electricity two costs were derived which correspond to the average plant load factors for the two scenarios. For coal based fuels two costs were calculated corresponding to coal costs of 15 and 30 £₁₉₇₅/tonne. the fuel costs can be converted to a primary energy cost by multiplying by the primary energy to fuel conversion efficiency. For ecample with coal at £30/tonne, gas costs 0.838 p/kWh or 2.33 £/GJ this is equivalent to 0.7*2.33 = 1.631 £/GJ of primary energy = 47.0 £/mtce. Of the total cost of gas 30/47 (64%) is due to the cost of coal and the remainder is due to capital, depreciation, labour and overhead charges.

Table 2.34 Fuel Cost Estimates for 2025 (1975 price levels)

		Electricity				Gas from		Liquids from
		Nuclear		Coal		Coal		Coal
Capital cost	(£/kWso)	500		150		80		60
Conversion efficiency	%	35		35		70		65
Load factor	%	77	60	19		50		80
Coal cost	£/tonne	-	-	15	30	15	30	15	30
Works cost	p/kWh
Capital	<10%-20yrs)	0.871	1.117	1.059	1.059	0.215	0.215	0.101	0.101
Fuel		0.400	0.400	0.536	1.071	0.268	0.536	0.288	0.577
Other		0.100	0.100	0.150	0.150	0.065	0.065	0.100	0.100
Ex-works cost	p/kWh	1.371	1.617	1.745	2.280	0.548	0.816	0.489	0.778
Other charges	p/kWh
Capital-transmission		0.450	0.450	0.343	0.343	0.040	0.040	-	-
Capital-refining		-	-	-	-	-	-	0.240	0.240
Overheads		0.250	0.250	0.250	0.250	0.250	0.250	0.150	0.150
Total fuel cost	p/kWh	2.071	2.317	2.338	2.873	0.838	1.106	0.879	1.168
	£/GJ	5.75	6.44	6.49	7.98	2.33	3.07	2.44	3.24
Primary energy   cost	£/mtce	58.0	64.9	65.4	80.4	47.0	61.9	48.7	60.7

source: after Chapman, Charlesworth and Baker (1976)
 

Primary Energy Demands and Costs

The primary energy demands for the two scenarios are shown in Table 2.35. Of the two, the electric vehicle scenario has the lower primary energy demand of 540 mtce, compared with 635 mtce for the liquid fuelled vehicle scenario. Also shown in Table 2.35 are the total costs of the energy. With coal at both 15 and 30 £/tonne the electric vehicle scenario costs substantially less than the liquid fuelled vehicle scenario.

Table 2.35 Total Annual Energy Demands and Costs

Scenario			Electric Vehicles			Liquid Fuelled Vehicles
			Primary energy	Cost		Primary energy	Cost
			mtce	£ million		mtce	£ million
Coal cost		£/tonne	-	15	30	-	15	30
Coal as	Coal		52	780	1560	52	780	1560
	Liquid		61	2788	3703	162	7403	9833
	Gas		145	6815	8976	-	-	-
	Electricity		-	-	-	44	2878	3538
	Total		258	10383	14239	258	11061	14931
Solar1			40	1880	2476	40	1880	2476
Primary Electricity			242	14036	14036	377	21871	21871
Total			540	26299	30751	635	34812	39278
Total Electricity			242	14036	14036	381	24749	25409

notes: 1 on a gas equivalent basis
 

Fuel Cost per Vehicle

There are four components to the cost of fuel for electric vehicles. These are:

1. Cost of electricity used to recharge vehicle battery whilst it is still in the vehicle (insitu recharging)
2. Cost of battery exchanges
3. Battery hire cost
4. Heating fuel cost.

The previous analysis of vehicle journey patterns was used to find the total number of battery exchanges on an average day. This was used to find the average number of exchanges required per vehicle. This was then used to find the number of kilometres which are run on insitu recharging. These calculations and the resultant costs are shown in Table 2.36.

Table 2.36 Electric Vehicle Annual Fuel Costs

Vehicle		Car	Van	Small Goods	Mid. Goods	Large Goods	Bus
Av annual   kilometres		15800	25500	27200	38700	55555	37777
Av number of   battery exchanges1		2	26	62	149	153	97
Cost per exchange2	£	1.96	2.42	4.04	6.32	14.86	6.96
Annual exchange   cost	£/yr	4	63	250	942	2274	675
Insitu recharge km3		15240	16140	14180	13370	18835	25167
Av fuel consumption	MJ/km	0.45	0.60	1.82	3.30	5.63	3.39
Insitu recharge cost4	£/yr	27	38	100	172	412	332
Battery hire cost5	£/yr	99	198	370	710	1964	699
Heating fuel	MJ/km	0.09	0.08	0.09	0.09	0.09	0.15
Heating cost6	£/yr	5	7	8	11	16	18
Total cost	£/yr	135	306	728	1835	4666	1724

notes:	1	Average number of battery exchanges per year  = 365 *  ("Average day" total number of exchanges, see Table 2.13) total number of exchanges
	2	From Table 2.19
	3	Insitu recharge km  =  Average annual km  -  number of exchanges  *  exchange range (see Table 2.9)
	4	Electricity at 1.4 p/kWh (3.89 £/GJ)
	5	It is assumed all batteries are hired rather than owned to remove the need for large transfer payments on exchanging a battery Battery hire costs are at 10 % (see Table 2.16)
	6	Liquid fuel at 3.24 £/GJ

The fuel cost for liquid fuelled vehicles are simply the product of annual range, fuel consumption and unit fuel cost. These are shown in Table 2.37. Also shown in this table are average annual fuel costs for each vehicle. These were found using the normalised populations (Table 2.5) as weighting factors.

Table 2.37 Liquid Fuelled Vehicle Fuel Costs

Vehicle group		Ann Av kms '000	fuel consum- ption	Fuel	Annual Fuel cost	norm- alized popul- ation	Av Annual Fuel cost
		km/yr	MJ/km	GJ/yr	£/yr		£/yr
Car   by   engine   size   litre	-1	15800	1.6	25.3	82	0.15	110
	1-1.5		2.0	31.6	102	0.50
	1.5-2		2.5	39.5	128	0.32
	2-3		3.2	50.6	164	0.02
	3+		3.8	60.0	195	0.01
Van   by   ULW   ton	-0.8	25500	1.8	45.9	149	0.25	184
	0.8-1		1.9	48.5	157	0.24
	1-1.5		2.6	66.3	215	0.51
Goods   by   ULW   ton	1.5-2	27200	4.1	112	361	0.10	466
	2-3	27200	5.8	158	511	0.23
	3-5	38700	6.7	259	840	0.10	840
	5-8	43000	10.0	430	1393	0.31	2244
	8+	71700	12.8	918	2974	0.25
Bus   by   seats	8-32	37777	5.8	219	710	0.12	1155
	32-48		8.5	321	1040	0.28
	48+		10.6	400	1297	0.60

Comparisons of fuel costs for vehicles run on the two fuels are shown in Figures 2.11 to 2.16.

Figure 2.11 Comparison of fuel costs for cars

Figure 2.12 Comparison of fuel costs for vans

Figure 2.13 Comparison of fuel costs for small goods

Figure 2.14 Comparison of fuel costs for middle goods

Figure 2.15 Comparison of fuel costs for large goods

Figure 2.16 Comparison of fuel costs for busses

2.4 Conclusions to the VRI study

The VRI study arose out of the earlier FTF study by Chapman, Charlesworth and Baker (1976) which examined the availability and suitability of the transport fuels most likely to be available to the UK over the next 50 years. The study re-examined the analyses contained in the earlier study and also examined the refuelling infrastructures required for an all electric road vehicle fleet or an all liquid fuelled road vehicle fleet. The electric vehicle infrastructure envisaged would allow electric vehicles to fill the same pattern of use as liquid fuelled vehicles.

In the VRI study it was shown that electric vehicles using a high energy density battery can be designed to have roughly comparable performance (top speed, acceleration, hill climbing ability etc.) with an internal combustion engine vehicle. It was found that the energy demands for heating electric vehicles, often neglected in analyses of this type, make little difference to the overall energy requirement for an electric vehicle (at least for use in the UK).

In contrast to the FTF study which considered only electric cars and vans, a comparison was made between a scenario in which the whole road vehicle fleet was electric with one in which all road vehicles were run on liquid fuels derived from coal. This further electrification reinforces the primary energy advantages of the electric vehicle scenario.

The study involved an investigation of two refuelling methods possible with electric vehicles, namely:

1. Insitu recharging - recharging the battery while it is still in the vehicle: which can be accomplished when and where the vehicle is garaged or parked
2. Battery exchange - exchanging a charged battery for a discharged one: which requires specialised equipment.

These two refuelling methods allow the major disadvantage (limited range) of electric vehicles to be overcome. The analysis of vehicle design and performance together with the data on vehicle use patterns lead to an estimate that approximately 2500 battery exchange stations would be required to meet the peak demand for batteries (August Saturday for cars and an October weekday for goods vehicles). It was assumed that exchange stations would be used by all road users. No consideration was given to what fraction of the vehicle fleet might be serviced by private battery exchange stations. Goods vehicle exchange played a large part in determining the size of the exchange system since they were responsible for the peak in demand for energy dispensed.

For the liquid fuelled vehicle scenario it was estimated that approximately 23,000 filling stations would be required.

It was found that without battery exchange (and so the need for hiring batteries) all electric vehicles had a lower fuel cost than the corresponding liquid fuelled vehicle. However with an average number of exchanges only the electric car had a lower fuel cost (including battery hire) than the corresponding liquid fuelled vehicle.

It is however worth questioning the wisdom of constraining the battery electric vehicle to follow the same pattern of use as today's liquid fuelled vehicles. In reality the vehicles themselves will generate their own use patterns. This may involve a renucleation of the industrial, shopping, and living areas in cities and towns and/or a mixed mode approach to long distance travel. Whatever does happen in the future it is almost impossible to predict with any degree of accuracy. Any approach to forecasting the exact patterns of use followed by electric vehicles in 2025 is likely to be much more subjective than the approach adopted in the VRI study. It is also worth noting that the only cost comparison made between the vehicles is of fuel cost. Also there was no consideration of alternative ways of overcoming the range limitations of electric vehicles such as the use of a mobile generator for use on long journeys (Lee and Corbett 1979).

Finally the systems effects associated with the widespread use of electric vehicles were considered. All the analyses were for a time in the future when it is assumed that the UK will be without indigenous natural oil or gas. The only fossil fuel available will be coal, which can be converted into a wide range of other fuels including synthetic natural gas and a liquid fuel for road transport.

In the liquid fuelled vehicle scenario the fixed demands for coal (that is demands which are not easily substitutable) are so large (250 Mtpa) that they equal the assumed maximum output of the UK coal industry. As a result all the flexible energy demands (ones which are easily substitutable), such as space heating, have to be met using electricity. In the electric vehicle scenario the fixed demands for coal were much smaller (in particular less liquid fuels and no coal for peak following electricity stations). As a result gas from coal could be used to meet all the flexible energy demands, giving a total coal demand of 250 Mtps. It is much more efficient in primary energy terms to use gas from coal for space heating than electricity. Consequently the electric vehicle scenario shows a very large saving in both primary energy and running cost over the liquid fuelled vehicle scenario. The electric vehicle scenario also required a much smaller installed electricity generating capacity and the load factor on this capacity was such that it could be totally nuclear. Consequently the systems effects are very much in favour of the electric vehicle.

2.5 Further Research

There are four main areas in which further research may be fruitful. These are the range requirements of vehicles, further developments of the computer models used, the development of more scenarios and practical work.

Range Requirements

The range identified as being important in the VRI study was the daily range of a vehicle since most vehicles are not used for at least eight hours overnight which is ample time to recharge batteries. There are two distributions of daily ranges which are of interest. The first (that for which estimates were made in the VRI study) is of the ranges of all vehicles (of a given type such as cars) on a particular day. This is a cross-sectional distribution. The other distribution is of the variation in daily ranges of individual cars over a period of time such as a year. This longitudinal distribution is of interest because it will indicate how often (if at all) and by how much the use of the individual vehicle exceeds its range.

Further work is required on the cross-sectional distributions of ranges since, as explained previously, the distributions derived for the VRI study (Figures 2.2 to 2.7) were extremely crude. A possible source of data for cars is the National Travel Survey which has been conducted by the Department of Transport every third year since 1972/3. The survey covers many aspects of household travel including the use of household cars. The results of the survey are held in a computer data base from which it is possible to extract tabulations of any desired variable. For goods vehicles a possible source is the Continuing Survey of Road Goods Transport conducted by the Department of Transport. Its results are also held in a computer data base from which tabulations can be extracted.

Deriving longitudinal distributions for individual vehicles is likely to be more problematical because surveys such as the National Travel Survey and the Continuing Survey of Toad Goods Vehicles are conducted over the year but individual returns are only for one week. Consequently the details know about any one vehicle are only for a week. As variations in daily ranges are likely to occur over a longer period than this, these surveys will be of limited use. However the National Travel Survey does record estimates of each cars annual mileage and wether it is a "first" car or a "second" car. Cross tabulations of maximum daily range in a week against average range in the week and annual mileage for individual months and "first" and "second" cars may be helpful in constructing synthetic longitudinal distributions. Other surveys of passenger travel, such as that conducted by British Rail for long distance journeys made by Manchester residents in 1974 (Beadle and Paulley 1979), may also help.

Developments to Computer Models

In the analysis two computer models were used. These were the driving cycle model used to determine electric vehicle energy consumption and ranges and the electricity system used to determine possible modes of operation of the electricity system, its load factor and required installed capacities.

There are two areas in which further work could be done on driving cycle models. The first is the further development of the model used. There are several ways in which the model could be improved. One of which is the introduction of a method to take account of the variation in battery discharge efficiency with power output. At present the model uses a set of efficiencies which remain constant over the whole driving cycle. Another possible area for development is to incorporate the operation of internal combustion engined vehicles. This would require the development of a model of the drive system (including the gear ratios and its rotational inertia) and the engine's performance/efficiency over its range of operation (power output and engine speed).

The second area in which work could be done on driving cycle models is in the modelling of the complete driver: vehicle: road system: traffic set of interactions. Work done by Nowottny and Hardman (1977) and further reported by Waters and Laker (1980) has modelled the driver: vehicle: road system set of interactions and was calibrated using the results of instrumenting vehicles to analyse their energy consumption (Easingwood-Wilson et. al. 1977). Such models could be very useful in design studies of vehicle energy use.

There are several areas in which further work on the electricity system model could be of benefit. These fall into three broad areas. These are refinement of the existing model, the introduction of more detail on road transport demands and the incorporation of intermittent sources of supply such as wave and wind power.

One of the short comings of the electricity system model is the use of a sinusoidal variation in average daily load over the year. Although the sinusoidal function used to model seasonal variations in demand may well be a reasonable reflection of the average load over a period of a month or so, it does not take account of the variation which occurs about this mean. There are two ways in which this could be overcome. One would be to introduce a random perturbation upon the sinusoidal function and the other would be to relate the variation in demand to weather variables such as temperature and sunshine. The latter approach is the more promising, particularly if combined with the other developments mentioned below. It is this approach which is being taken by Barret (1981).

There are several ways in which the detail about road transport used in the model could be improved. These include the incorporation of daily traffic flows. The reports on the 50-point traffic census (Tanner and Scott 1962, Dunn 1962,3,4,5,6,7, Dunn and Shepard 1968, Dunn 1970,1,2,3,4) give estimates of total vehicle miles on all classified roads in Great Britain over the years 1961 to 1972. These could be combined with the results of the monthly road side surveys reported in Highway Statistics (Department of the Environment annual) to obtain estimates of vehicle kilometres by cars, vans, rigid goods, articulated goods and busses. These results could be reduced to a common basis so that they could be used in projections in which the total annual mileages were different, and could then be used in place of the Fourier patterns used in the VRI study.

A second set of details which could be included are those on when vehicles are not in use during the day. These could be estimated from studies on vehicle flow rates, such as Gyenes (1973). The incorporation of such details would require modification to the model's consideration of "logic loads" so that it did not use more load than was connected at any one time.

Finally it would be advantageous to incorporate variable sources of supply, such as wind and wave power, into the model. An early attempt to do this was for wave power (Vimukta et al 1978) had the short coming that it used randomly genrated hourly data to represent wave power availability. Although the distribution of hourly outputs corresponded to that over a time period of several years there was no attempt made to model the auto correlation of the power available. This was due to lack of basic data. Until recently it has not been possible to use this approach to investigate wavepower and storage in anything but a speculative fashion as done by Chapman (1977). However a recent paper by Winter (1980) indicates that the Meteorological Office have developed a wave forecasting model which could be used to obtain time series estimates of wave spectra at 30 minute intervals for a large number of offshore sites. There is also better data available on wind power and Low is currently developing the model to incorporate-weather-record based estimates of Windpower availability (Low 1981).

Scenarios

There are two ways in which the scenarios used in the VRI study could be developed. The first would be to make simple alterations to the two existing scenarios to generate modified scenarios. Such modifications might involve taking the electric vehicles scenario and have all goods vehicles run on liquid fuels and have a separate operator run a battery exchange system for buses so that public battery exchange would only be used for cars and vans. Another modification would be similar to this but would involve the use of mobile generators for use on long journeys instead of battery exchange.

The second way in which the scenarios used could be developed would be to use more extensive and sophisticated scenarios of energy use in the remainder of the economy such as those in A Low Energy Strategy for the United Kingdom (Leach et. al. 1979) and Energy Technologies for the United Kingdom: An appraisal for R, D & D planning (Department of Energy 1979). The use of such scenarios would also enable the dynamics of introducing synthetic liquid fuels from coal and of electric vehicles over time between now and 2025 to be examined.

Practical

Finally two areas which were identified in the VRI study and which require research, and possibly then development, are a cheap and simple 40 amp plug and socket, and a solid state recharger. One of the things which would be advantageous in such a recharger would be a variable charging voltage so that the current delivered could be kept constant for a large part of the charging cycle unlike the simple charger design envisaged in the VRI study.

2.6 Lessons I learnt from the VRI study

As with the Freight Transport Statistics study I learnt several valuable lessons from the refuelling infrastructure study.

Conformity

As with the FTF study there were several ways in which the VRI study was constrained to conform with the work and views of others. This ranged from technical issues such as electric vehicle energy consumption where modifications were made to the electric vehicle model until the results it produced were within a believable range. This believable range was set by the work of others in the field and their expectations. At the other end of the scale the GDP, energy and transport projections which were used were chosen on the basis that they would not prejudice the results of the study rather than on the basis that we thought they were particularly likely.

Support for the energy demand projections used in the FTF study was taken from them being broadly in line with those of the Department of Energy. It was the FTF demand projections which were used in the VRI study. However by the time it was conducted in 1978 and subsequently revised in 1980 general expectations of energy demand had fallen considerably, as can be seen from Figure 2.17.

Figure 2.17 Forecasts of Primary Energy Demand in 2000

In both the FTF and VRI studies Department of the Environment forecasts of traffic were used, not because they were thought to be particularly good, that is what is likely to happen, but because their use would not prejudice the studies results in the eyes of government.

Cooking the books

In the process of revising the VRI study several small changes were made in the starting assumptions and projections. For example the recalculation of the vehicle energy consumptions and the "exchange" ranges led to an increase from 200km range to 280km range for cars. In its turn this led to a reduction from 1 million to 260 thousand for the number of exchanges required on an August Saturday. This is an example of the effects of changing highly sensitive variables in an analysis. Although it was not done in this case such changes can often be made without seriously affecting the plausibility of the starting assumptions but radically change the end result. It is often very easy to cook the books.

Changes in behaviour

In the VRI study one of the difficulties for which no adequate solution could be found was that of new technologies changing patterns of behaviours. This came to light when considering the introduction of electric vehicles. The original intention was to make a comparison between scenarios in which the vehicles performed the same journey patterns. However it was found that the costs of journeys would, on a per kilometre basis, be much higher for long journeys than for short journeys. This would be likely to have the effect of changing behaviour patterns so as to reduce the number of long journeys made. No way of modelling such changes in behaviour were immediately apparent and even if they had been their use would have rendered the two scenarios incompatible.

Reasonableness of constraints

There were other ways in which the construction of comperable scenarios led to problems. For example in the liquid fuelled vehicle scenario there limit on coal production of 258 million tonnes of coal seems much less reasonable than in the electric vehicle scenario. This is becuase in the liquid fuelled case there would be much greater pressures to increase coal production than in the elctric case.

Data

Yet again there were problems with the data upon which models were constructed. For example the data on vehicle ranges is very sparse to the extent that the analysis carried out was more wishful thinking than anything else. Another example was the lack of data on seasonal variations in electricity demand for the electricity demand model.

Summary of lessons learnt

From these lessons I learnt that forecasting studies and scenario construction is often tailored to be acceptable to its audience. This entails keeping within the general bounds of current expectations. I also confirmed my belief that it would be possible to produce any desired result by the judicious choice of starting assumptions. Another thing I found out was that changes in behaviour can be ignored, but if they are, the results will be that much less believable or useful. Another thing was the "comparable" scenarios are not necessarily very useful because the constraints of comparability can impose unrealistic constraints. Finally I again found that data can cause serious problems.

[1] I have yet to understand how liquid fuels can be derived from electricity stored in batteries.

About | Contents
Previous: 1. Freight Transport Statistics | This: 2. Vehicle Refuelling Infrastructure | Next: 3. Transport Input-Output Study
Copyright © Michael Baker 1981,2005. All Rights Reserved.