Energy and land use

But before we embrace land use controls as a primary device to throttle the flow of energy in urban areas, or even as a useful supplement to other conservation strategies, we should exa

Energy and land use

An instrument of US conservation policy?

D a l e L K e y e s

Energy consumption in society is a

function, among other things, of its

spatial distribution The author

describes studies of energy usage

in metropolitan areas of the USA,

simulation studies of alternative

building types and the potential for

energy conservation Rearranging

the urban housing pattern into

more compact high-rise patterns

could lead to significant savings in

domestic and transport energy con-

sumption Some policy of en-

couraging more energy conserving

urban s t r u c t u r e s m i g h t be

beneficial but caution is needed in

case the detrimental side-effects

outweigh the benefits

The author is Research Associate

with the Urban Institute, 2 1 0 0 M St

NW, Washington DC, USA

1 Ford Foundation Energy Policy Project,

A Time to Chose: America's Energy

Future, Ballinger Publishing Co, Cam-

bridge, Massachusetts, 1974; Federal

Energy Administration, Project

Independence Report, Washington, DC,

US Government Printing Office,

November 1974

z See, for example, Anthony Downs, 'S-

queezing Spread City', New York Times,

March 17, 1974; pp 38-47; Herbert M

Franklin, 'Will the New Consciousness of

Energy and Environment Create an

Imploding Metropolis?' A m e r i c a n

Institute of Architects Journal, August

1974, pp 28-36

3In recognition of this need the US

government is sponsoring research on

land use-energy linkages, some of which

is being undertaken by the Urban

Institute However, we need not wait for

this work to bear fruit in order to take

stock of what is currently known

The past two years have witnessed a groundswell of interest in energy conservation issues and a number of studies on the relative merits of various conservation options Two studies in particular have detailed the expected benefits and to a lesser extent the costs of selected conservation strategies available to federal, state, and local governments in the U S A ? These strategies range from fuel price deregulation to various types of mandated conservation measures Partly because these studies did not investigate the entire set of available options and partly because none of the strategies is completely painless, attention has increasingly focused on avenues to energy conservation yet unexplored Among these is 'land use', its popularity due in no small part to its imprecise definition Land use means so many different things to so many people that few can deny its potential for affecting some aspect of the way energy is used This interest in reducing energy consumption by altering land use patterns

is evidenced by a spate of articles appearing in the popular press and various non-technical journals} Compact, well planned communities are more energy efficient, it is argued, and governments should encourage - if not require - urban growth which is consonant with sound planning principles But before we embrace land use controls

as a primary device to throttle the flow of energy in urban areas, or even as a useful supplement to other conservation strategies, we should examine the magnitude of suspected energy savings associated with various land use patterns and the costs or difficulties involved in manipulating these patterns)

U r b a n spatial structure The overall relationship between patterns of land use and energy consumption (holding constant the influence of climate, population characteristics, intensity of industrialisation, end-use efflciencies and the like) derives from two intermediate relationships, one involving building types and the other travel behaviour It is more illuminating

to speak of the spatial structure of urban areas rather than the pattern

of land uses, for the energy used in any city is presumably a result of its three dimensional structure Theoretically, where densities are high, residential and commercial units are frequently smaller and often share common walls or are located in large structures,

Energy and land use

~To determine the magnitude of the

effect on space heating, and cooling one

needs only to characterise spatial struc-

ture in terms of population densities or,

more precisely, the mix of building types

But in order to specify the quantitative

relationship between urban spatial struc-

ture and transportation energy use, one

must be able to characterise development

patterns along several dimensions - the

relatige location of activity centres, the

nature of the transport system, and the

geography of the population density func-

tion In addition, the degree of local in-

tegration of different land uses and the

travel consequences must be separated

from the impact of region-wide land use

integration That is, the impact of a high

degree of metropolitan-wide segregation

created by regional shopping c e n t r e s -

few in number but large in size - may be

obviated by the presence of small com-

mercial establishments in close proximity

to many residential areas

6 Regional Plan Association Inc and

Resources for the Future, Regional

Energy Consumption, Regional Plan

Association, RPA Bulletin 121, January

1974

6 Space heating and cooling efficiencies

associated with high densities should

theoretically apply to the industrial and

commercial sectors as well as to the

residential sector However, consumption

data is typically available only for total

energy use rather than just for space con-

ditioning Since energy used for purposes

other than heating and cooling is of

greater importance in the commercial and

especially in the industrial sectors, the

discussion here is limited to residential

energy use (Space heating and cooling

represents approximately 7 0 % of the total

energy used in homes)

providing economies for space heating and cooling; where employment and commercial centres are located close to residential areas, travel will be reduced; and where development is located along transportation corridors and at sufficient density, people and goods can be transported by more energy efficient modes

In order to measure the differences in energy use as a result of variations in urban spatial structure, a relatively detailed description

of both energy consumption patterns and spatial structure is required for several urban areas which display significant structural differences 4 Additional information on differences in energy use among different types of buildings can be used to make inferences about the heating/cooling effects of urban structure where direct data are unavailable Unfortunately, disaggregated data on energy use in urban areas is almost totally lacking, while description of the spatial structure found in various cities rests almost exclusively on population densities and density gradients Likewise, detailed empirical studies of energy consumption for individual buildings are not numerous Simulations of both hypothetical and real urban structures and building types must therefore be used to supplement and

in some cases replace, empirical analyses of observed consumption and spatial structure or building characteristics

Energy studies Perhaps the most detailed study of metropolitan energy use in the USA has been conducted for the New York City region, where energy use for each of the region's 31 counties has been estimated by economic sector and by fuel type 5 Urban spatial structure, however,

is only represented by two variables - gross population density and the extent of the transit system As is shown in Table 1, the New York Metropolitan Area (which is composed of nine of the 31 counties in the region) has the highest area-wide and central city population densities and the most extensive transit service Of the five most populous metropolitan areas (and in fact of any area) in the USA One might therefore expect that the energy consumed in both the transportation and residentiaP sectors would be lower for New York than for any other area or for the country as a whole, once climate and other factors are accounted for Table 2 shows that the

Table 1 Population and transit characteristics of the five largest US metropolitan areas

(1970 and 1972 data respectively)

Central city b Vehicle miles Vehicle miles of SMSA population of transit

population density (people per year SMSA a (x 106) (people p~r per sq (miles

milex 10 ) milex 103 ) x 106) c

Los Angeles/

transit operated per year per capita (Miles per person )

a Standard Metropolitan Statistical Area

as defined by the US Office of Management and the Budget

b The central city within the SMSA as defined by the US Office of Management and the Budget

c Approximate figures

Source: The population data are taken from US Bureau of the Census, US Census Population, 1970, Vol 1, Part A,

1970 The transit data came from American Transit Association, 1972 Transit Operating Report, ATA, no date

Table 2 Annual residential and trans-

portation energy consumption in the New

York City area and in the USA (in million

BTU per person)

Net per capita energy use a

Residential Transportation

Greater

b

a Net energy use is the total quantities of

coal, oil, gas, steam and electricity in BTU

equivalents delivered to each section

(Electricity is converted at the direct heat

equivalent of 3 4 1 2 Btus per kWh.)

b This is approximately equivalent to the

NYC SMSA

Source: Joel Darmstadter, Conserving

Energy, Prospects and Opportunities in

the New York Region Baltimore: The

Johns Hopkins Press, 1975, Table 1-1

7 Joel Darmstadter, op cit Note b, Table 2

These are money rather than total per-

sonal income figures

a Reported income elasticities for expen-

ditures on electricity in the USA range

from about 0 3 to 0.7 Thus, a 1% in-

crease in income would lead to a 0 3 to

0 7 % increase in expenditures Although

these elasticities are based on total per-

sonal income (which is slightly larger than

money income) and apply strictly to elec-

tricity only, the relatively low values imply

that the unexpectedly high consumption

levels in New York cannot be totally ex-

plained by an income argument For a dis-

cussion of i n c o m e - e x p e n d i t u r e

elasticities, see Robert Halvorsen,

'Residential demand for electric energy',

Review of Economics and Statistics, Vol

57, pp 12-18, (1975)

Since gross residential density is a poor

surrogate for building types, we have

attempted to statistically relate the RPA

data on average per capita, per unit and

per square foot energy consumption in

each county to the percentage of one-

family units and the percentage of

buildings with 5 or more units found in

each county In this way, we were able to

show that an increase in the percentage

of one-family units does contribute to in-

creased consumption, although per capita

income, the price of the various fuels, and

other factors appear to be much more im-

portant in determining consumption

levels The results of this regression

analysis are roughly comparable to those

of the energy simulations undertaken by

Hittman Associates, to be discussed sub-

sequently That is, the largest structures

seem to be most efficient on a per unit

basis but not on a per capita basis

I ° A comprehensive discussion can be

found in Melvyn D Cheslow and Mary

Lou O l s s o n , Transportation and

Metropolitan Development, The Urban

Institute, Working Paper No 5 0 4 9 - 0 7 ,

1975

Energy and land use

average energy used for transportation by a New York resident is indeed less than that used by persons residing elsewhere in the country However, the per capita residential consumption is 15%

greater in the New York area than the average for the country as a whole This rather puzzling observation may be at least partially due

to the relatively higher per capita income level found in the New York area ($3911 versus $3118 in 1969 dollars for the greater New York City area and the USA, respectively 7) However, this 26% increase in income level could not sufficiently offset the 15% increase in consumption to explain the difference between the observed and expected results 8 Climate may be a more important factor but its impact cannot be assessed without additional information

In order to gain further insight into the observed relationships between urban spatial structure and energy use in the New York region, intra-regional variations can be probed Again, descriptions of spatial structure are limited, this time exclusively to population density The reported results are essentially a reaffirmation of the previous findings: residential consumption expressed on either a per capita or per square foot basis is approximately the same for each county in the region (Density ranges from 53 persons per square mile

in Sullivan County to almost 70 000 in Manhattan.) 9 Intra-regional density-transportation energy linkages are some- what less ambiguous Per capita transportation energy use decreases

at an increasing rate up to about 4 0 0 0 0 persons per square mile, and then displays an increase as density increases still further These results could be partially due to the effects of congestion at very high population densities, but can also be attributed to the method of computation If employment or daytime county populations were used as the basis for computing per capita consumption, the reversal of the density-consumption relationship would be moderated if not eliminated

The data on energy use in the New York City region thus provide suggestive evidence in support of the hypothesised transportation energy-spatial structure relationships and largely inconclusive evidence for the proposed residential energy-spatial structure relationship But the evidence is limited and subject to problems of data collection and analysis In addition, no information is provided

on the effects or urban structure other than population density However, these preliminary attempts to probe the residential consumption data more thoroughly suggest that the use of data on building type rather than population density may provide more insights, and furthermore, that multi-family structures do appear to be less thermally efficient than one-family units (See Reference 9) Although few researchers have investigated metropolitan energy consumption patterns directly, metropolitan transportation research

is impressive, at least in volume if not in terms of consistent approaches and definitive findings? ° Travel behaviour in urban areas

is typically characterised by the number of trips taken by individual residents, the travel mode used, and the length and speeds of those trips These characteristics are affected by a host of factors, many of which have little to do with urban structure In fact, several studies have shown that household characteristics are much more important than structural features of urban areas H Unfortunately, not all studies have accounted for household characteristics when estimating the effects of spatial structure

Energy and land use

l~Stowers and Kanwit, for example,

observed that family size, income, and life

cycle stage are more important than den-

sity of the neighbourhood or distance

from the central business district in deter-

mining trip generation rates See J.P

Stowers and E.C Kanwit, 'The use of

behavioral surveys in forecasting

transportation requirements', Highway

Research Record, Highway Research

Board, 1966, pp 44-51

~2 See, for example, William Smith and

Associates, Patterns of Car Ownership,

Trip Generation, and Trip Sharing in Ur-

banizedAreas, prepared for the Bureau of

Public Roads, US Department of

Transportation; also Harold D

Deutschman and Nathan L Jaschik, 'In-

come and related transportation and land-

use planning implications', Highway

Research Record, Highway Research

Board, 1968, pp 52-65

~3 See, for example, Wilbur Smith and

Associates, op cit and Council on

Municipal Performance, 'City perfor-

mance', Municipal Performance Report,

V o l l N o 6 , 1 9 7 5 , p p 1 1 - 1 7

14See, for example, David M Branston,

'Urban traffic speeds- I: a comparison of

proposed expressions relating journey

speed to distance from town centre',

Transportation Science, February 1974,

pp 35-49

~5 See Allan M Voorhees and Associate,

Factors and Trends in Trip Lengths,

NCHRP Report No 48, Highway

Research Board, 1968; also Automobile

Manufacturers Association, 1972

Automobile Facts and Figures, AMA, (no

date)

~6 R.W Anderson, Residential Energy

Consumption, Single Family Housing,

Hittman Associates Inc, March 1973 and

M Tokmanhekin and D.G Harvey,

Residential Energy Consumption,

Multifamily Housing Final Report, Hitt-

man Associates Inc, June 1974

t~ A.D Little Inc, Residential and

Commercial Energy Use Patterns, 1970-

90, Project Independence, US Federal

Energy Administration, November 1974

They used a relatively simple estimation

method (known as the 'Degree-Day

Method') developed by the American

Society of Heating, Refrigeration, Air

Conditioning and Ventilation

Although comparison of findings from the various transportation studies is inhibited by the lack of comensurate units of analysis, the following can be stated as general findings:

• As population density increases within urban areas, the number of non-pedestrian trips decreases This appears to

be true for both generally high and generally low density urban areas 12

• As both population and employment density increases within urban areas, the percentage of trips taken by automobile decreases ]3

• As distance from the central business district increases in British cities (a surrogate for decreasing population density), the average speed of traffic at any point increases ~4

• Average metropolitan trip length has not been shown to vary consistently with average metropolitan population density, although some differences between central city and suburban lengths have been observed ~5

It is obvious that the existing empirical evidence does little more than whet the appetite Some findings (but not all) lend support to the hypothesised relationship between transportation energy use and urban spatial structure, but more is needed to substantiate it even in a qualitative sense

Simulation studies o f alternative building types Several models have recently been developed to estimate the heating and cooling loads of various types of buildings They relate energy requirements to design and orientation factors, construction methods, construction and insulation materials and characteristics of the heating/cooling equipment One such study, on the energy requirements of protypical single and multi-family housing units found in the Baltimore, Maryland-Washington, DC area, concluded that sizable differences in energy requirements among these units are attributable to the type of structure within which the unit is housed.l~ The results appear in Table 3 A general trend toward reduced energy loads with increased density of units is observed, due both to the inherent thermal efficiency gains and the smaller size of multi-family units On the basis of area the difference between single family detached units and units in highrise structures is about 20-30% Although these results apply, strictly speaking, only for the Baltimore- Washington region, the relative energy requirements among the various units should be applicable to most areas in the county since the climate in this region imposes both heating and cooling demands The Arthur D Little Organization has explicitly incorporated regional climatic variations in a similar study of residential energy consumption 17 They estimated a heating and cooling load differential between prototypical single family detached and high rise units of about 65% for a given area, averaged across all regions

However, the error involved in both of these sets of estimates is largely unknown (although the Hittman analysis was more rigorous) and the values should be accepted as very rough approximations of

Energy and land use

Table 3 Average annual energy requirements for heating and cooling prototypical housing units in the Baltimore-Washington area a

In-structure energy consumption (Therms)

of floor area Single-family detached 1000-1100 0 5 9 0 - 0 6 4 8 250 275

a The value for the single-family units are not entirely comparable with those for the multifamily units as type of

construction and building materials were not completely standardised A range of

values is indicated for the single-family

unit with masonry walls as the prototype tested differed somewhat in insulation characteristics from the multifamily units

Source: R.W Anderson, op cit and M

Tokmahekin and D.G Harvey, op cit

Reference 16

energy savings due to building type In order to judge better the validity of these studies, careful analyses of actual residential consumption patterns as a function of household and building characteristics (including type) is urgently needed

18 Jerry B Schneide and Joseph R Beck,

Reducing the Travel Requirements of the

American City: An Investigation of Alter-

native Urban Spatial Structures, Research

Report No 73-1, Departments of Urban

Planning and Civil Engineering, University

of Washington, August 1973

19 The authors found, for example, that by

rearranging the location of most people

and jobs in the area, a 66% reduction in

total travel (total person-minutes for work

trips) and a 87% reduction in the max-

imum link load or level of use on the most

highly used highway could be achieved

Another arrangement involving a

rearrangement of 'only' a quarter of the

total population and jobs would produce a

43% and a 63% reduction in these fac-

tors, respectively The results indicate that

travel requirements can be minimised by

locating most people and jobs at the least

central node and by matching the number

of people and jobs at each node However

the optimal arrangements for Seattle

would require extensive relocation of the

existing population and employment base

20 Jerry L Edwards and Joseph L

S c h o f e r , Relationships B e t w e e n

Transportation Energy Consumption and

Urban Structure: Results of Simulation

Studies, Department of Civil Engineering,

Northwestern University, January 1975

Studies o f alternative urban structures

We now turn our attention to studies of alternative arrangements of urban areas These studies estimate either total travel from which transportation energy can be calculated or energy used for travel directly

In the first study, undoubtably the most theoretical of those to be reviewed, Schneider and Beck have developed a procedure for searching through a very large number of possible spatial arrangements of jobs, residences, and transportation networks ~8 Using simple assumptions about travel behaviour (based primarily on the gravity model), those spatial arrangements are then selected which minimise total travel time or maximise other objectives In order to lend an air of realism to the undertaking, the urban area in and around Seattle, Washington, was characterised by a very simple network of transportation nodes (employment and residential centres) and links (highways) and the impact of redistributing population and employment among the nodes was estimated ~9

Methological restrictions severely limit the general application of their findings For example, only work trips made by automobile were considered, while the translation of total travel time and link loading factors into energy savings requires more specific information about vehicle speeds and flow rates But perhaps the most restricting characteristic of the analysis is the simplistic representation of urban structure Still, the results shed some light on the range of travel savings which may be achievable, ignoring for the moment real world constraints to urban development processes

Another study compares the expected transportation energy savings from a more limited number of urban structural alternatives 2° However, each alternative is characterised in much more detail than

in the Schneider and Beck study and those spatial structures which are used for test purposes are intended to be better approximations of feasible urban development patterns

Energy and land use

21 M o r e specifically, the elasticities of

consumption w i t h respect to the second

m o m e n t s of both population and employ-

ment about a central point are con-

siderably less than one

zz Eric Hirst and John C Moyers, 'Efficien-

cy of energy use in the United States',

Science, Vol 179, 1973, pp 1 2 9 9 - 1 3 0 3

Examples of energy-intensive as well as energy-efficient spatial structures were created from three basic urban shapes (concentric ring, linear, and polynucleated) and several combinations thereof A spatial distribution pattern of basic employment was specified together with a transportation network including up to three nodes, each with unique service characteristics Total population, total employment, labour force participation rates and certain travel behaviour characteristics of the residents were drawn from an existing city - Souix Falls, South Dakota - and the spatial distribution of residential and commercial activities specified by the Lowry land use model In this way 37 different combinations of spatial structures and transportation network structures were constructed Total travel levels by mode were calculated using simple gravity and route assignment models Finally, energy consumption was computed from information on the length and speed of all types of trips

The results of the 37 simulations indicate that the energy consumption levels for the most and least energy-intensive hypo- thetical urban areas differ by a factor of about ten However, one must take care to sort out the effects of urban structure alone (shape, population distribution and employment distribution) from those of the purely transportation-related factors (available modes and service characteristics) This is made difficult both from the interactive nature

of urban structure and transportation and from the way the results are presented in the report Still, some disaggregation can be made, if only in an approximate manner

If those test cities which differ only in shape are compared, then the linear cities appear to consume least and the concentric ring cities most, perhaps by as much as a factor of two However, 'shape' is in many ways a surrogate for population and employment distribution The authors report that energy consumption is negatively correlated with employment and population concentration; that is, those structures with compact spatial distributions are more efficient However, relatively large increases in concentration have to take place to achieve significant savings 2~

Looking at variations in the transportation system alone, an 80- 85% decrease in energy consumption is predicted for a concentric ring city which realises an increase in transit usage from 0 to 70% But this figure must be considered highly approximate, as the urban structural variables were not held completely constant in these comparisons In addition, other estimates of modal shift impacts have been less optimistic Hirst and Moyers, for example, report a potential decrease in energy use of about 25% for an increase in transit patronage from the current level o f 3% to a future level of 51%.22

We can again conclude that dramatic changes in urban structure will produce sizable savings in the amount of energy used to transport people However, since the range of structures was determined a

priori, we cannot be sure that each of the test structures is feasible, given the past and probable future processes which lead to urban development This is true despite the use of a partially validated land use model to distribute secondary employment and population in a realistic manner once the location of the primary employment had been specified Moreover, it is difficult to glean from these results implications for development added over the next few decades to existing urban areas of various structural types - the issue which is of most interest from a policy perspective

230 ENERGY POLICY September 1976

23 Margaret Fulton Fels and Michael J

Munson, 'Energy Thrift in Urban

Transportation: Options for the Future', in

Robert H Williams, ed., The Energy

Conservation Papers, Ballinger Publishing

Company, Cambridge, Mass 1975

2"Two other options examined by Fels

and Munson are based primarily on

changes in urban structure and are

worthy of further examination

One option assumes that people will

retain their desire for low density living

but will locate in order to shorten their

journey to work Employers wilt likewise

respond by clustering in a few large

employment centres throughout the

metropolitan area Low density residential

villages would thus develop around these

centres, greatly reducing the journey to

work for the average resident On the

other hand, non-work trips are likely to be

unaffected, or, if commercial activities are

also clustered in a few large centres, will

probably be longer Significant net

savings are nevertheless achieved; ap-

proximately 20% and 35% savings are es-

timated, relative to the most energy con-

suming option for 1985 and 2000,

respectively The savings from this

development trend which appear in the

report incorporate several technological

and energy price changes as well The

savings noted here were calculated from

the reported data with the effects of other

changes factored out

Another option emphasises high den-

sity residential living but assumes a non-

contiguous pattern of employment Non-

work travel is thus greatly reduced while

average work trips are lengthened But

the former effects outweigh the latter pro-

ducing approximately 25% and 30%

savings in consumption, with respect to

the highest energy patterns, by 1985 and

2000, respectively

zs James S Roberts, Energy, Land Use,

and Growth Policy: Implications for

Metropolitan Washington, Metropolitan

Washington Council of Governments,

June 1975

Energy and land use

The final two studies confront the relevant issues more directly Both examine the energy impact of future growth options for specific metropolitan areas and both assess the relative effectiveness of alternative but related conservation options

The first is a study of urban transportation energy consumption in the Trenton, New Jersey, Metropolitan Area between 1975 and the end of this century33 Among the many conservation options considered, some relate directly to urban spatial structure (eg, increasing home-work proximity) while trends in development patterns underlie or condition the use of other strategies Realism in the projected effects is improved by using the population, employment, transport network, and transportation behaviour characteristic of Trenton as the basis for the estimates Future changes in these characteristics must still be assumed, but the sensitivity of the results to these assumptions can be measured

Fels and Munson calculate that a four fold and ten fold difference

in per capita energy consumed for transportation in Trenton is possible, in the extreme, by the years 1985 and 2000, respectively This

is the difference between the 'everybody wants a luxury car' option (where long and more frequent trips are taken by richer people living further apart and driving less efficient autos) and the 'energy consciousness' option (where fewer and shorter trips are taken by people living closer to each other and to work and travelling by more efficient modes and in more efficient autos) Between these extremes are options which feature innovative forms of mass transit, fuel price increases, technologically modified autos, and various combinations

of these 24 The final study investigates the energy implications of alternative growth scenarios for the Washington, DC area 2s The scenarios selected for examination are broadly representative of different urban structures They include (a) S p r a w l - low density, non-contiguous residential growth at the fringe with new employment primarily at the metropolitan centre, (b) W e d g e s a n d c o r r i d o r s - all new development located in a radial configuration along rapid rail transit routes and concentrated near stations, (c) B e l t w a y o r i e n t e d - all new develop- ment located along a cirumferential highway, and (d) D e n s e c e n t r e -

higher density, more concentrated development focused on the metropolitan centre

The energy implications for future travel were examined using a simple transportation model calibrated to explain current travel behaviour in the area The results show that the sprawl pattern would

be about twice as energy-intensive (considering automobile consumption only) as the dense centre and transit oriented patterns

by 1992 Again one cannot say that a 50% savings is realistically obtainable, only that this is the difference between these two extremes

of spatial structure, given the assumptions about travel behaviour upon which the simulation is based

Before proceeding it is important to place the simulation studies just discussed in perspective Most of the alternative urban structures simulated are grossly unrealistic given current US development patterns Even where conscious efforts have been made to provide residential opportunities in close proximity to employment centres, a very poor matching of employees and households has resulted In order for many of the simulated structures to be realised a considerable change in attitudes and preferences must occur

Energy and land use

2s This assumes that the same number of

people would be accommodated in units

of the same aggregate floor space This

may not be totally realistic as some

reduction in floor space typically accom-

panies a move from a single family

detached unit to a multi-family unit (Floor

space is usually traded for location)

Some additional savings may be expected

simply from a reduction in the size of the

average unit found in high density

developments

Planning is certainly not a sufficient condition to alter development patterns

Relative effectiveness of a land use policy Relative comparisons among alternatives assume that the absolute value for each is known In the case of energy conservation strategies, the available evidence will only allow us to reduce the degree of speculation about the effectiveness of each With this caveat in mind

an attempt will now be made to estimate the savings that could reasonably be expected to obtain from the control of new development patterns and to place these in the context of savings potentially achievable from alternative conservation strategies

Looking first at energy consumed for residential space heating and cooling, we have three pieces of evidence from which to derive an estimate of probable savings The two simulations predict that a 20- 30% (Hittman Associates) or 65% savings (Arthur D Little) could be achieved if all new development resulted in the construction of high-rise buildings rather than single family detached units 26 A more modest estimate is justified since even in very high density growth patterns, a mix of housing types will be found In addition, the absence of empirical verification of the simulation results, as judged by the New York City findings, further argues for a reduced estimate Thus, a value of 20% for the potential savings accruing to new high density development will be used for discussion purposes

Probable transportation energy savings from efficient development patterns can likewise be estimated from the simulation studies, modified by the empirical evidence Table 4 shows the energy differences between the most energy consuming development patterns, and other more efficient patterns tested in the two studies which simulated the effects of new growth to existing urban areas The most efficient scenarios are not presented as their realisation seems most unlikely even under extreme conditions This is probably true for the dense centre pattern in the Washington study as well The

Table 4 Simulated transportation energy use differences for various patterns of incremental, growth

Total additional energy consumption (from the base year a) and percent reduction in this increment from the base case

(in BTU per year x 1012 )

(Total) (%)

Luxury car (base case) 8.5 - Home-neighbour

proximity (modifiedb) 6.6 22

Home-work proximity

Sprawl (base case) Wedges and corridors Dense centre

(Total

1992 (%)

70.6 59.5 16 35.1 50

2000

(Total) (%) 16.8 10.5 37

a Early 1 9 7 0 ' s for the first three patterns, 1 9 7 3 for the last three patterns

b These options have been modified from the original study by assuming no technological modification of the auto, no introduction of innovative transit modes,

no use restrictions on the auto, and no

increase in fuel prices

Source: The first three patterns are taken from the Trenton, NJ study (Fels and

M u n s o n , op cit Reference 23) and the last three from the Washington, DC study (Roberts, op cit Reference 25)

Energy and land use

Table 5 Probable reduction in total US energy consumption per year from efficient development patterns (%)

Percent of Savings (%) Estimated reduction

(1968) Savings (%)

Urban residential space

Urban residential air

Urban automobile and

a Stanford Research Institute, Patterns

of Energy Consumption in the United States, prepared for the Office of Science and Technology, Executive Office of the President, 1972; US Bureau of the Census, US Census of Population, 1970,

Vol 1, Part A, 1970; and Eric Hirst,

Energy Intensiveness of Passenger and Freight Transportation Modes, 1950-70,

ORNL-NSF-EP-44, Oak Ridge National Laboratory, April, 1975

2~ The adoption of a comprehensive plan

however, does not insure that local

development will actually take place in

accordance with the plan Local zoning

ordinances need only reflect the plan's

general intent and, in any case, are

susceptible to significant change through

the issuance of variances

other development options represent less extreme deviations from what may be expected in the absence of more stringent land use control In fact, the wedges and corridors scenario is based on the region-wide land use plan accepted by all local governments in the Washington area 27 The reductions in incremental energy consumption over the base cases which are associated with these growth patterns are thus much more realistic To some degree they may still be over-statements due to possible exaggerations inherent in the base cases (ie, neither base case appears to be a simple extrapolation from present trends), but they provide a basis for arriving at reasonable approximations Again, the lack of strong empirical evidence to substantiate the simulated savings lends credence to this approach

Since some but not all aspects of the home-neighbour and home- work proximity patterns could likely be accomplished by vigorous growth management efforts and since the wedges and corridor growth concept was practical enough at least to appear in the Washington area comprehensive plan, a value of 20% will be used for probable savings by 1985 resulting from energy efficient changes to existing urban structures

As a first order approximation of the magnitude of savings which could be achieved at the national level by 1985, the estimated urban structure efficiencies can be converted into a fractional reduction in annual US consumption levels and applied to the projected annual consumption level in 1985 Table 5 illustrates the calculation of the fractional reduction in annual US consumption using 1968 as the base year This saving should be applied only to the total new consumption represented by future growth Using the US Federal Energy Administration's (FEA) conservative projection for 1985, energy consumption of 103 quadrillion British Thermal Units (BTU) per year, the difference between projected 1985 and actual 1972 consumption is 31 quadrillion BTU per year 28 Assuming that the distribution of fuel consumption among economic sectors and the distribution of people between rural and urban areas remains approximately constant, the total one year savings in 1985 would be approximately 0.9 quadrillion BTU Other conservation actions have

Energy and land use

2 e Federal Energy Administration, Project

Independence Report, Federal Energy Ad-

ministration, November 1974, p 26

29 Federal Energy Administration, op cit

Reference 28

30 Fels and Munson, op cit Reference 23

31 See William K Reilly, The Use of Land:

A Citizen's Policy Guide to Urban Growth,

Thomas Y Crowell Co, New York, 1973

and Fred Bosselman, David Callies, and

John Banta, The Taking Issue, Govern-

ment Printing Office, 1973, Washington,

DC

32 The recent defeat of a national land use

bill in Congress is evidence of public sen-

timent against governmental encroach-

ment on private property rights, even

though the bill had land use planning,

rather than regulation, objectives For a

more detailed discussion of suggested

changes in land use policy and the im-

plications for private property rights, see

Peter G Grown, The American Law

Institute Model Land Development Code,

The Taking Issue, and Private Property

Rights, The Urban Institute, URI 13300,

1975

been assessed by FEA and the savings from those that are most closely related to land use consideration are as follows 29 (in 1012 BTU):

Mandatory fuel efficiency standards for autos (20 mpg) 1.9 Special gasoline taxes 1.2 Mandatory thermal efficiency standards for new buildings 1.0 Seen in this context, the likely savings from more efficient land development patterns are significant but not dramatic

The relative effectiveness of still other related policies can be anticipated from the results of a study already discussed From data reported by Fels and Munson, it appears that increases in fuel prices combined with reductions in auto weight and modification of the internal combustion engine will produce savings about three times greater than those from land use measures alone) °

Of course, many of these conservation measures are highly interrelated and should not be considered alternatives in a strict sense Thermal efficiency standards are likely to affect the most inefficient building types (single family detached dwellings) to the greatest extent, reducing the energy differential between high and low density development patterns Likewise, imposition of automotive fuel efficiency standards will serve to reduce the differential between compact and dispersed development And fuel price increases in general will create an impetus toward more energy efficient urban structure The effectiveness of a land use policy is therefore really a function of which other policies will also be employed

'Costs' of a land use policy

In judging the suitability of any policy, relative efficacy is but one of two considerations Cost, acceptance, or general feasibility is the other We are far from being able to say with any authority what the monetary effects are of applying land use controls to create a given spatial structure Certainly, limiting the supply of land for certain uses will affect land prices which, in turn, will be reflected in the cost of urban goods and services But little more than that assertion can be made The 'costs' of a land use policy can more reasonably be discussed in terms of legal and political acceptability

The right to private ownership and use of land in the United States has a rather substantial history, though public bodies (especially local governments) have made significant inroads in regulating land use during this century Recently, arguments for a much greater public involvement have been proffered by groups anxious to prevent 'land abuse' and environmental degradation 31 These groups espouse public regulation of land use far in excess of what has traditionally been allowed by the courts No longer would development controls be fettered by community health, safety, and morals criteria alone The new powers would be based on an expanded and rather amorphous definition of 'community welfare' Energy conservation would undoubtably be considered a justifiable community objective thus legitimising alterations in urban spatial structures to achieve this end, often without compensation for those whose property would be reduced in value or even confiscated But such an expansion of public powers is far from being universally accepted by the courts and has hardly met with unanimous public acceptance 32 So the question

234 ENERGY POLICY September 1976