energy demand in the norwegian building stock: scenarios ... · energy demand in the norwegian...

ARTICLE IN PRESS

Energy Policy 37 (2009) 1614–1627

Contents lists available at ScienceDirect

Energy Policy

0301-42

doi:10.1

� Corr

0314 Os

E-m

journal homepage: www.elsevier.com/locate/enpol

Energy demand in the Norwegian building stock: Scenarioson potential reduction

Igor Sartori a,�, Bjørn Jensen Wachenfeldt b, Anne Grete Hestnes a

a Department of Architectural Design, History and Technology, Norwegian University of Science and Technology (NTNU), 7491 Trondheim, Norwayb SINTEF Building and Infrastructure, 7465 Trondheim, Norway

a r t i c l e i n f o

Article history:

Received 23 May 2008

Accepted 16 December 2008Available online 12 February 2009

Keywords:

Energy demand

Building stock

Scenario analysis

15/$ - see front matter & 2009 Elsevier Ltd. A

016/j.enpol.2008.12.031

esponding author at: SINTEF Building and Inf

lo, Norway. Tel.: +47 22 96 55 41; fax: +47 22

ail address: [email protected] (I. Sartori).

a b s t r a c t

A model has been developed for studying the effect of three hypothetical approaches in reducing

electricity and energy demand in the Norwegian building stock: wide diffusion of thermal carriers, heat

pumps and conservation measures, respectively. Combinations of these are also considered. The model

has a demand side perspective, considers both residential and service sectors, and calculates energy

flows from net to delivered energy. Energy demand is given by the product of activity and intensity

matrices. The activity levels are defined for the stock and the new construction, renovation and

demolition flows. The intensity properties are defined in archetypes, and are the result of different

energy class and heating carriers share options. The scenarios are shaped by combining the activity

flows with different archetypes. The results show that adopting conservation measures on a large scale

does allow reducing both electricity and total energy demand from present day levels while the building

stock keeps growing. The results also highlight the importance of making a clear distinction between

the assumptions on intensity and activity levels.

& 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Energy demand in the building stock in Norway representsabout 40% of the final energy consumption, of which 22% goes tothe residential sector and 18% to the non-residential sector.Similar figures are reported for the EU countries (DGET, 2004) andare also valid for western countries in general. In Norway, wherethere is a strong dependency on electricity for heating purposes,electricity covers about 80% of the energy demand in buildings.These numbers reflect the potential for improvements inherent inthe building stock. Policies targeting a reduction of energy andelectricity demand in the building sector should be based onsound models of the stock and the mechanisms inside it that areultimately responsible for the energy demand. Nonetheless, someaspects seem not to be fully understood, and this may represent asignificant handicap for policy making.

In his work Myhre (2000) makes a thorough analysis of theresidential sector in Norway and explores different scenarios forfuture energy demand until 2030. He considers a number of agegroups and a number of retrofit options for the different agegroups. However, the assumptions on the amount of renovationactivity seem not to be based upon empirical evidence. Further-

ll rights reserved.

rastructure, Pb 124 Blindern,

69 94 38.

more, possible substitution of electricity with other carriers is notconsidered.

A scenario analysis is also performed by Johansson et al. (2006,2007) for the building stock in a Swedish region, comprising boththe residential and non-residential sectors. These studies take awide perspective on the energy consumption related to thebuilding sector; they analyze the energy flows starting from thebuildings’ demand and up through the supply chain to finallycalculate numbers on primary energy and associated CO2 emis-sions. This is surely a worthwhile approach, and the estimation ofCO2 emission is often a most important goal because of thenational reduction targets set by the Kyoto protocol. Nevertheless,while these studies embrace in their analysis the entire energychain, they seem to put the emphasis on the supply side ratherthan on the demand side. The assumptions on how changes mayhappen seem to be based solely on the economic lifetime ofheating systems, and the estimation of possible energy demandreduction appears to overlook the inertia that characterizeschanges in the building stock. This may have lead to too generousassumptions about the possibility for decreasing energy demand,or at least about its timing.

In relation to this, an interesting work is that of Nassen andHolmberg (2005) on the Swedish building sector. They make anhistorical analysis of the building stock and the related energydemand in the period 1975–2000, questioning why the totalenergy use has remained almost unchanged – actually slightlyincreased – despite the great potential for improvement already

www.sciencedirect.com/science/journal/jepo

www.elsevier.com/locate/enpol

dx.doi.org/10.1016/j.enpol.2008.12.031

mailto:[email protected]

ARTICLE IN PRESS

I. Sartori et al. / Energy Policy 37 (2009) 1614–1627 1615

known at the time. Nassen and Holmberg (2005) refer to Steenet al. (1981) who had illustrated scenarios for energy demand in2000 being less than one third of the energy demand in 1975,already achievable with ‘‘present known best technologies’’. Theiranalysis concludes that because the most prominent measuretaken to reduce oil dependence after the oil crises was anincreased supply of electricity – through development of nuclearpower plants – the potential of improving energy efficiency waspoorly utilized. They warn that in a coming transition it will beimportant to better balance the attention between supplysubstitution and energy efficiency.

In this work the authors explore energy scenarios in thebuilding stock from a demand side perspective. The net energydemand is set as the starting point for the analysis. Bothresidential and service sectors are included in the analysis. Energyclasses and heating carriers share are used to describe the energyintensity of buildings. In parallel, the amount of floor area relatedto construction, renovation and demolition activities is estimated.Intensity and activity data are combined into a scenario analysisaimed at evaluating the potential for electricity and energydemand reduction. The performed analysis has its boundaries‘‘at the door of the building’’, leaving out any consideration thatgoes beyond the delivered energy. An analysis of the correspond-ing primary energy and CO2 emissions (possibly with otherscenarios that consider supply side options) could be performedin a supply side study, when taking delivered energy figurespresented here as a starting point. This is, at any rate, out of thescope of this work. The purpose of this work is to estimate thepotential for reducing the demand of energy, especially electricity,in the Norwegian building stock.

1.1. Definitions

The energy needed in a building in order to satisfy specific enduses is named net energy, as shown in Fig. 1, and it comprises:heating, cooling, ventilation, hot water, lighting and other electricequipment. Technical systems are used to meet such demand, andthe totality of energy delivered to the building site in order tooperate all the technical systems is named delivered energy.

The delivered energy is the object of this study. The deliveredenergy comes in the form of different energy carriers, which can begrouped in two main groups: electricity and thermal carriers. Theon-site renewable energy is considered as free energy in the sensethat it is already available to the end user without the necessity tobe delivered by any carrier. The calculations performed in thiswork go from the net energy demand to the delivered energy.

1.2. Heating systems in the Norwegian building stock

The energy system in Norway is characterized by a peculiarelectricity production that is for more than 99% based onhydropower (DGET, 2004). The strong hydropower developmentof the last half century has guaranteed abundant and cheap

Technicalsystems

Net energy

On-site renewables“free energy”

Transformationand

Distributiondelivered energy primary energy

Calculation direction

Fig. 1. Sketch of the energy chain.

electricity to the country; electricity has also been accompaniedby the image of being environmentally friendly because it is basedon a renewable source. These aspects have had a strong influencein shaping the energy demand in Norwegian dwellings (Bøeng,2005) and in all buildings in general.

While the demand continues to grow both in the industrial andthe building sectors, the further potential for large scale hydro-power development is limited due to protection of the remainingnatural waterfalls. As a consequence national electricity con-sumption has increased more than the national power production,and unlike before Norway has actually been a net importer ofelectricity in the Scandinavian NordPool market six of the tenyears from 1996–2005 (NVE, 2008). Therefore, there is a stronginterest in the opportunities to reduce electricity dependency, andthe building sector could play an important role.

The strong dependency on electricity in the Norwegianbuilding stock can be seen when looking at the heating systemsinstalled. According to the last census (Statistics Norway, 2001)about 70% of dwellings are equipped with space heaters and/orfloor cables for direct use of electricity, either as the only systemor in combination with wood or oil stoves. Only 12% of dwellingshave a hydronic system. In total, as much as 93% of respondentsstate that they use (also) electricity for heating their dwellings.Nevertheless, 70% of dwellings are equipped with more than oneheating system, the most popular being direct electricitycombined with a wood stove (38%). This means that they cancombine electricity with other carriers, at least to cover part of theheating need. The situation is less dramatic in the non-residentialbuildings, but figures on electricity use are still high. According todata reported by Enova (2006) on about 1900 buildings monitored(of which approx. 1800 non-residential) at least 62% of the non-residential buildings have a central heating system, either alone orcombined with direct electricity space heaters. Nevertheless,electricity is used in 67% of the cases as one of the fuels alsowhere there is a central hydronic system.

2. Data sources

Energy demand can be expressed as the product of activity andintensity. The term activity means the amount of square meters offloor area, and the term intensity means the energy demanded persquare meter. Eq. (1) reports this relation with the appropriateunits

EnergyTWh

y

� �¼ Activity½m2� � Intensity

kWh

m2 y

� �(1)

where Energy, Activity and Intensity are matrices whose structureis described later. The Intensity matrix is used in two ways: inanalysing historical data it is the output resulting by dividing themeasured energy consumption by the measured floor areaactivity. In the scenario analysis the Intensity matrix is an inputthat multiplies the expected activity in order to forecast theenergy demand.

2.1. Activity

A number of sources are available on the building stock inNorway, but unfortunately not all the information available issuitable for the purpose of this work. The various sources areanalysed in detail in the reports Sartori (2006) and Sartori andWachenfeldt (2007). The reports conclude that the only suitablesource of data is given by the GAB register (Grunneiendom-,Adresse- og Bygningsregisteret), the computer register containinginformation about ground properties and addresses in Norway.The GAB register contains information on both residential and

ARTICLE IN PRESS

Table 1Data on buildings by year 2005.

Category GAB code Share (%) Floor area (m2)a

Residential sector 100 320,829,869

Farm, single- and row-house 11x, 12x, 13x 80 255,069,051

Apartments block 14x, 15x 12 40,044,225

Holiday house 16x 8 25,716,593

Excluded from the model

Residential garage and similar 17x, 18x, 19x

Industry, agriculture and fishery 2xx

Service sector 100 127,116,947

Office and shop 3xx 39 49,192,584

Transport and communication 4xx 5 6,836,984

Hotel and restaurant 5xx 7 8,757,735

Education, culture, sport, religious 6xx 27 33,906,419

Hospital and nursery 7xx 7 8,805,852

Prison and emergency preparedness 8xx 2 2,466,283

Other 9xx 13 17,151,089

a Gross floor area.

Residential sector, Area & Delivered Energy

0

60

120

180

240

300

360

1975

Are

a [ m

illio

n m

2 ]

0

10

20

30

40

50

60

Ene

rgy

[ TW

h/y

]

Service sector, Area & Delivered Energy

0

30

60

90

120

150

180

Are

a [ m

illio

n m

2 ]

0

10

20

30

40

50

60

Ene

rgy

[ TW

h/y

]

electricity district heating wood gas oil free heat HP floor area

1980 1985 1990 1995 2000 2005

1975 1980 1985 1990 1995 2000 2005

Fig. 2. Historical data on area and delivered energy for: (a) the residential sector,

and (b) the service sector.

I. Sartori et al. / Energy Policy 37 (2009) 1614–16271616

non-residential buildings, is based on a complete census ofbuildings, and contains coherent data on the gross area. The onlydrawback is that the electronic version was established in 1983.

Every building is registered in GAB and labelled using a 3-digits code to identify the building categories. The data in GABneed to be read, elaborated and grouped in categories by softwarein order to produce useful information; such software has beendeveloped by the authors and the results have been comparedwith other sources, when possible, in order to validate the tool(Sartori, 2006). Table 1 reports the GAB coding to an aggregatedlevel – one or two digits – and the corresponding floor area in year2005.

The data from GAB had to be aggregated because data onenergy consumption are not available at the same level of detail.Two sectors are defined: the residential sector, comprising allcategories of residential buildings, and the service sector, compris-ing all other categories. A few categories were excluded from themodel, either because they have no (or negligible) energyconsumption or because their energy consumption is studiedseparately from the building sector, i.e. being part of the industrialsector. It is worth noticing that the residential sector accounts for71% of the total floor area.

2.2. Energy

The main information on energy consumption comes fromStatistics Norway (2007a) that provides data on all the energycarriers from year 1976 until 2005. The statistics provide data forelectricity, district heating (from 1991), wood, gas and oil;negligible quantities of charcoal and coke are also reported butdisregarded here. The aggregation level is such to allow only adistinction between energy consumption in the residential sectorand in the service sector. In Statistics Norway (2007a) the servicesector properly defined does not include the item ‘‘other uses’’,which is instead included in the DGET (2004) in the so called non-residential sector. For this reason the service sector as presentedhere is responsible for approx. 12.5% of Norway’s final energyconsumption, and not 18% as stated in the introduction for thenon-residential sector.

Additional information was needed to properly include theeffect of heat pumps. This technology is spreading rapidly in

Norway, and considering that this trend may continue in thecoming years it becomes important to estimate how much heatpumps contribute to satisfy the buildings’ net energy demand.Based on the studies by Eggen (2005) and Grorud et al. (2007),Wachenfeldt and Sartori (2007) estimated the electricity con-sumption and the heat supplied by heat pumps in standardizedaverage operating conditions. In order to asses the averagecoefficient of performances (COPs), it was supposed the mostcommon situation where air-to-air heat pumps are installed inresidential buildings, while other types are installed in non-residential buildings.

Merging the data from statistics on energy carriers and theestimations on heat pumps results in the data series repre-sented graphically in Fig. 2. The term free heat HP is usedto indicate the difference between the heat actually suppliedto the building and the electricity required to run the heatpump. The free heat HP is therefore not part of the deliveredenergy.

As the stock has been growing nearly constantly so has theenergy demand, both in the residential and the service sectors. Inyear 2005, the residential sector represents 71% of the total floorarea and is responsible for 64% of the total energy consumption. Itis also clear that electricity is the prevailing energy carrier in bothsectors. In the residential sector wood has become the largestthermal carrier as the use of oil has slowly decreased. Districtheating has found only a marginal diffusion and the gas share is sosmall as to be barely visible in the graph. The contribution of heatpumps is recent and remains small even though rapidly growing.In the service sector oil seems to be phasing off more slowly whilewood is not important at all. District heating has grownconsiderably in the last two decades while the gas share hasremained marginal. Here too heat pumps contribution is stillsmall but rapidly growing.

ARTICLE IN PRESS


2.3. Intensity

The sources of information used to gather data on the intensityare the reports from Pettersen et al. (2005) and Wigenstad et al.(2005), subsequently modified by Thyholt et al. (2007). Theseprovided the basis for the revised building code (TEK, 2007).Pettersen et al. (2005) and Wigenstad et al. (2005) describe amethodology that should constitute the basis for the Norwegianclassification of buildings based on their energy performan-ce—the Norwegian implementation of the European Directiveon Energy Performance of Buildings (EPBD, 2002 and accompany-ing standards). These two documents will be referred to as theEPBD proposal.

The proposed methodology is based on both the evaluation ofaverage performance for the present stock, named stock referenceRs, and the prescribed performance for new and largely renovatedbuildings, named regulation reference Rr. The values of Rs and Rr

are specified for a number of building categories; for newconstruction indicative numbers are available also for thedifferent end uses, calculated under standardized referenceconditions (Thyholt et al., 2007). Estimations for holiday houseconsumption are taken from Wachenfeldt (2004). The values arereported in Table A1 in Appendix A.

These data are given for the Oslo climate. Nevertheless,harmonizing the data on regional distribution of dwellings(Statistics Norway, 2001) with the climatic zones defined byEnova (2006) based on the Degree-Days method, it can beestimated that about 76% of dwellings are located in areas witha climate comparable to or milder than the Oslo climate. Hence,adopting the Oslo climate as an average climate valid for theentire building stock can be considered as a reasonable approx-imation.

The values of Rs and Rr are central to the definition of theenergy classes given in the EPBD, as described later. The values forRs and Rr refer to heated floor area, while the data from the GABregister refer to gross floor area. Hence, in order to calibrate themeasured intensity data (Intensity ¼ Energy/Activity) with thereference values Rs, it is necessary to use a conversion factor forthe reference floor area. The conversion factor is obtained bymaking the measured intensity equal the Rs intensity. Doing so iscomparable to studying an equivalent building stock placed in theOslo climate, and whose intensities correspond perfectly with theestimations made in the EPBD proposal. The resulting conversionfactors from gross to heated floor area are: 0.69 for the residentialsector, and 0.80 for the service sector.

Fig. 3 shows the historical development of intensity per heatedfloor area for both the residential and service sectors. The share ofelectricity is also shown.

Intensity & elec

0

100

200

300

400

500

1980

Inte

nsity

per

hea

ted

area

[k

Wh/

m2 y

]

Residential, intensity Service, intensity

1985 1990 1995

Fig. 3. Historical data on intensity and electricity

Concerning the residential sector the intensity seems to haveslightly decreased over the last decade. The observed improve-ment can be linked to two causes. The first and possibly strongerreason is that in recent years, since the beginning of the ‘90ies,there has been a drop in the construction of single- and double-family houses and row-houses, accompanied by an increasedconstruction of dwellings in blocks of flats (Statistics Norway,2007b). Apartment dwellings have better energy performancesthan dwellings in detached houses, see Table A1 in AppendixA. The second cause may be that the stricter regulations inthe building code may have gradually had an effect (previousbuilding codes in Norway are from 1987 and 1997). Theconstruction of new buildings that are more and more energyefficient, together with the simultaneous demolition of some oldbuildings, has resulted in a decrease in the overall averageintensity of the stock.

Concerning the service sector, the energy intensity does notseem to decrease in the same way as for the residential sector. Onthe contrary, observations from Enova (2006) would rathersuggest that new buildings may consume more than older onesdespite the stricter building code’s prescriptions. Part of thereason may be the increased ventilation rates and increased use ofglazing in the facades of new buildings compared to older ones.

Finally, it shall be noticed that Rs values refer to deliveredenergy because that is the actual measurable quantity.The regulation Rr values, instead, refer to net energy. Therelation between net and delivered energy, for each carrier, isgiven in Eq. (2)

Enet ¼ Edelivered Zsys (2)

where Zsys is the overall efficiency (or coefficient of performance,COP) of the technical system, including distribution and controlefficiencies. Pettersen et al. (2005) and Wigenstad et al. (2005)present an extensive evaluation of the system efficiencies fortypical installations in both the residential and the service sector.The overall average efficiencies for each energy carrier arereported in Table A2 in Appendix A.

3. Method

One basic assumption made here is that a consistent andenduring change in the net energy demand of a building can beachieved only when a building undergoes major renovation, i.e.extensive works that can change the thermal performance of itsenvelope. Thus a building, or rather a generic square meter, isrepresented by a certain energy class from the moment it is builtuntil it is eventually renovated.

tricity share

0 %

20 %

40 %

60 %

80 %

100 %

elec

trici

ty s

hare

[%]

Residential, electricity Service, electricity

2000 2005 2010

share for the residential and service sectors.

ARTICLE IN PRESS


The share of heating carriers is allowed to change regardless of(major) renovation activity. Where a hydronic heating system isavailable the shift from one carrier to another is normally notproblematic. On the other hand, if a dwelling or an office does nothave a hydronic system, the installation of pipes, radiators and soon may hamper the adoption of thermal carriers. In any case, theobserved shift from electricity toward thermal carriers in theresidential sector is quite limited and it is mainly toward wood,see Table 3. A larger shift is observed for the service sector, wheremost buildings are already equipped with hydronic heatingsystems (62%). Therefore, it is reasonable to assume that theexpected shift toward thermal carriers can take place indepen-dently from (major) renovation work.

The combination of energy class and heating carriers shareform an archetype. Every generic square meter of floor area in themodel is represented by one archetype or another. In addition tothe existing stock, three flows of floor area are defined: newconstruction, demolition and renovation. Each year the newconstruction flow is assigned to some archetypes (or fractions ofthe total flow assigned to different archetypes) according to thescenario’s assumptions. The demolition flow is simply removedfrom the stock. The renovation flow re-circulates in the stock,meaning that a portion of the existing floor area ‘‘migrates’’ fromsome archetypes to others, in accordance with the scenario’sassumptions.

The initial features of the stock are given by the average valuesfor the period from 1996 to 2005, which is the most recent knowndecade of data. The model is then run for a simulation period thatgoes from 2006 to 2035.

3.1. Energy classes

The energy classes are labelled with letters from A to G, whereA means most efficient and G means least efficient. TheNorwegian scale proposed in Pettersen et al. (2005) andWigenstad et al. (2005) is modified with respect to the Europeanscale (EN 15217, 2007); both scales are reported in Table 2.

In the Norwegian scale the class E is supposed to represent theaverage for the existing stock (Rs), while class C is the regulationrequirement for new buildings (Rr). The class A represents the bestperforming buildings and is similar, for example, to the ‘‘passivehouse’’ standard established by the Passiv Haus Institut inGermany (residential only).

Table 2Definition of energy classes.

Class EN 15217 scale Modified Norwegian scale

A p0.5 �Rr p0.5 �Rr

B pRr p0.75 �Rr

C p0.5 � (Rr+Rs) pRr

D pRs p0.5 � (Rr+Rs)

E p1.5 �Rs pRs

F p2.0 �Rs p1.5 �Rs

G 42.0 �Rs 41.5 �Rs

Table 3Heating carriers share.

Year Electricity direct (%) District hea

Residential Average 1996–2005 80.7 0.9

Trend to 2035 61.8 2.7

Service Average 1996–2005 71.7 6.5

Trend to 2035 31.6 29.2

3.2. Heating carriers share

As mentioned, in Norway electricity is the single mostimportant carrier used for heating purposes. Nevertheless,electricity is often complemented by other carriers, and the actualuse of electricity or alternative carriers depends on how thesystem is run (i.e. dependent on price variations).

The heating carriers share refers to the net energy demand forheating, and so it expresses the preference given by the users toalternative energy carriers used to satisfy their heating needs. Theheating carriers share is derived by combining information ondelivered energy, net energy demand and system efficiencies. Thevalues observed in the period 1996–2005 were averaged. Thetrend observed in the period 1996–2005 is continued linearlyuntil year 2035, and both average and trend values are reported inTable 3. For reasons of simplicity also the estimated heat fromheat pump is treated like an energy carrier. The term ‘‘electricitydirect’’ does not include the electricity for driving heat pumps.

Significant changes can be noticed between the initial and thefinal values, especially in the service sector. If the observed trendwill continue, direct use of electricity will be more than halved by2035. District heating and heat from heat pumps will becomenearly as important as direct use of electricity. The use of gas willalso increase, while oil is almost phased off. In the residentialsector the direct use of electricity keeps being the most significantcarrier, but with a smaller share than in the starting period. Theuse of wood and heat pumps increase considerably, while gas anddistrict heating keep playing a marginal role. Oil use is completelyphased off by 2035.

3.3. Archetypes

An archetype is ‘‘a statistical composite of the features foundwithin a category of buildings in the stock’’ (ECBCS, 2004). Thestructure of an archetype is summarized in Table 4.

On the left side of the table is the net energy, subdivided intothe energy needs: electric, cooling and heating. The electric needis the sum of electricity demanded for fans and pumps, lightingand other electrical equipment. Cooling means both space coolingand cooling of ventilation air; it is assumed that cooling isprovided by electricity-driven systems. The heating need is thesum of space-, ventilation- and tap water heating. The total netenergy demand is uniquely determined by the building’s energyclass.

On the right end side of the table is the delivered energy,subdivided into the energy carriers. The free heat from heatpumps is reported in brackets because it is not an actual carrier,and it is not included in the total of delivered energy. The figuresfor delivered energy are obtained by the net energy consideringthe heating carriers share, see Table 3, and the carrier efficiencies,see Table A2 in Appendix A.

A closer look at the archetype structure is given in Appendix B.The archetypes shown are representative of stock and regulationreferences, Rs and Rr, when adopting the observed average heatingcarriers share, see Table 3. The archetypes describing Rs and Rr

ting (%) Wood (%) Gas (%) Oil (%) Heat from HP (%)

8.9 0.3 8.1 1.1

18.7 2.5 0.0 14.4

0.1 0.9 15.4 5.5

1.2 7.3 3.0 27.7

ARTICLE IN PRESS

Table 4The archetype’s structure.

Archetype XYZ: sector X, energy class Y, heating carriers share Z

Net energy Delivered energy

Energy need (kW h/m2y) Energy carrier (kW h/m2y)

Total net ‘‘-Heating carriers share

and efficiencies -’’

Total delivered

Electric x1 Electricity y1

Cooling x2 District heating y2

Heating x3 Wood y3

Gas y4

Oil y5

(Free heat HP) (y0)

Energy class dependent Heating carriers share dependent


represent the energy classes E and C, respectively. It shall benoticed that in the residential sector heating needs are as high as78% in class E, and 61% in class C. In the service sector heatingneeds are at 61% in class E, and 45% in class C. Electric and coolingneeds are the same in class E and C; the only difference in theresidential sector is given by the adoption of mechanicalventilation in new buildings. In class B and A energy needs are75% and 50% of class C values, respectively, as defined in Table 2.In the service sector delivered energy results being lower than thecorresponding net energy demand. This is due to the COP of heatpumps and cooling machines.

Archetypes were created for all heating carriers share options:the two values reported in Table 3 plus two other alternativesspecified by the scenarios. Combining four heating carriers shareswith five energy classes (from A to E) and two sectors adds up to atotal of 40 archetypes. A further level of detailing may includesubdivisions based on climatic zones, type of area (urban/rural,high/low energy density), disaggregation of building sectors intocategories, age groups and types of heating system available.Apart from the difficulties in resourcing the necessary information– unfortunately often scarce or inexistent – the number ofarchetypes would grow rapidly, and the potential benefit comingfrom a more detailed description of the stock should be weightedagainst the increased complexity of the data structure.

4. Scenarios

The scenario analysis investigates the potential of threedifferent approaches for reducing total energy and electricitydemand in the Norwegian building stock. The three approachesare: wide diffusion of thermal carriers, wide diffusion of heatpumps and wide diffusion of energy conservation measures. Theenergy conservation approach is combined with the two others toform additional scenarios. The scenarios analysed in this work arethe following:

�
Base: reference scenario, based on observed trends. � Thermal: scenario achieving wide diffusion of thermal carriers. � Heat Pump: scenario achieving wide diffusion of heat pumps. � + Conservation: scenarios achieving wide diffusion of energy
conservation measures, in addition to the other assumptions inall three previous scenarios.

The scenarios are defined by the Activity and Intensity matrices.The Intensity matrix is defined by the assumptions on energyclass and heating carriers share, and is therefore different for eachscenario. The Activity matrix is defined by the assumptions onfloor area, and is therefore common to all scenarios. Nevertheless,there is an exception. As there is high uncertainty on the estimates

of the renovation flow, three levels are considered: low, middleand high. All three levels are applied to the scenarios in order toestimate the model’s sensitivity to this parameter.

4.1. Assumptions on activity

The basic assumptions about the activity levels are presentedin Table 5, where all data are presented in gross floor area. Data onstock and new construction are taken from the GAB register. Thestock increase in both residential and service sectors appears to benearly linear, see Fig. 2. This is in line with the population increasefor the same period, which also has been nearly linear (StatisticsNorway, 2007c). The same statistics projects the expectedincrease in Norwegian population in the next decades to continuewith approximately the same pace. Hence, the stock is alsoexpected to continue its linear growth, and the flow of newconstruction is set approximately equal to that of the observedperiod. The average construction activity between 1996 and 2005corresponds to about 1.0% and 1.4% of the 2005 reference stock, inthe residential and service sectors, respectively. This indicates thatthere has been a relatively higher activity in the service sector.

Regarding demolition, data had to be inferred from othersources, and the scarce information available was mainly forresidential buildings. According to Myhre (2000) and Rødsethet al. (1997) a reasonable estimate would be to assume that 5000dwellings are demolished every year. With an average size ofdwellings of about 130 m2, as from GAB database, this results in650,000 m2/y of demolition activity. This demolition rate corre-sponds approximately to 0.2% of the residential reference stock ofyear 2005. For the service sector, in absence of better guesses, thedemolition rate is also set to 0.2% of the 2005 reference stock,which means 250,000 m2/y.

Regarding renovation, data availability is even scarcer than fordemolition. In their work Sartori et al. (2008) analyze the availablesources and compare them with the results of their model (adynamic Material Flow Analysis, MFA, applied to Norway’sdwelling stock). Even though their results seems to be anunderestimate compared to other sources, their conclusion isthat in the coming decades renovation is likely to overtake newconstruction as the major activity. In their medium scenario theyestimate that 19% of the 2000’s dwelling stock had been renovatedin the past 30 years. This would mean about 2 millions m2/y ofgross floor area, expected to rise up to 4 millions m2/y in the nexttwo/three decades. However, due to the high uncertainty in theestimates, three levels are considered here for the renovationactivity. They are set to 2, 3 and 4 millions m2/y for the low,middle and high level, respectively, in the residential sector. In theservice sector, in a similar way, the three levels are set to 1, 1.5 and2 millions m2/y. All three levels are applied to the scenario inorder to estimate the model’s sensitivity to this parameter.

ARTICLE IN PRESS

Table 5Scenario assumptions on Activity.

Residential sector Service sector

(m2/year) Notes (m2/year) Notes

Observed data, 1996–2005

New ffi3,280,000 GAB register, average ffi1,740,000 GAB register, average

Demolition ffi650,000 0.2% of 2005’s stock

Renovation ffi2–4,000,000 Sartori et al. (2008)

Scenarios data, 2006–2035

New 3,000,000 1,500,000

Demolition 650,000 250,000 0.2% of 2005’s stock

Renovation 2–3–4,000,000 Low–middle–high level 1–1.5–2,000,000 Low-middle-high level


Considering the middle level of renovation, the assumptionssummarised in Table 5 lead to the following conclusions. In theresidential sector, 6% of the reference 2005 stock will bedemolished by 2035 and 66% will remain unchanged. Inthe service sector, 6% of the reference 2005 stock will bedemolished by 2035 and 59% will remain unchanged.

4.2. Assumptions on energy class

As a consequence of the observations made in the paragraphIntensity, residential and service sectors have different initialsituations. In the residential sector there is a slight improvementof the stock performance, while in the service sector the situationis approximately unchanged. The assumptions are summarised asfollows:

�
Residential sector:The reference stock in 2005 is in class E since this is the classthat is supposed to represent the average of the stockaccording to the EPBD proposal. It may be argued thatrepresenting the entire stock with a single energy class is acoarse approximation and more age groups should be used. Onthe other hand, as many buildings have been renovated itwould be misleading to make a categorisation strictly based on‘‘as built’’ characteristics.Rather, the stock is supposed to be normally distributedaround the central value Rs ¼ 190 kW h/m2y, and with nooverlap with the regulation value Rr ¼ 121 kW h/m2y for newconstruction,1 see Tables B1 and B2 in Appendix B. Thebuildings that are demolished and renovated yearly aresupposed to be those with worst energy performance. As aconsequence, the remaining stock will have a slightly betteraverage performance. In 2035 the remaining unchanged 66% ofthe stock (the remaining part of the normal distribution ‘‘bell’’)will have an average performance of 177 kW h/m2y instead ofthe original 190 kW h/m2y. This is named the class E*.Renovation in 2005 already achieves the energy class E*, andnew construction in 2005 is in class D. � Service sector:
The unchanged stock is supposed to remain in class E inaverage. Renovation and new construction in 2005 are also inclass E.
� Base, Thermal and Heat Pump scenarios: the above assumptions
are valid until 2035.
� All+Conservation scenarios: all scenarios start from the basis of
the EPBD, which produces its effects from 2010 with newconstruction in class C and renovation achieving class C as

1 This means choosing a standard deviation s ¼ (Rs�Rr)/3.

well. This is called the initial state. Then, a transition period isset for from 2010 to 2020. At the end of the transition period allnew buildings have class A and all renovated ones reach classB. This is called the final state. During the transition periodthere is a gradual shift from the initial to the final state, withmore and more new and renovated buildings that conform tothe final state.

The transition period is needed to simulate the time required forextensive diffusion of certain design concepts, materials, technol-ogies and practical expertise. The final state is assumed in class Bfor the renovation flow (and not in class A) because technicaldifficulties and costs are generally higher in renovation than innew construction.

4.3. Assumptions on heating carriers share

Where not otherwise specified, the assumptions hold true forboth the residential and service sectors. They are summarised asfollows:

�
The original stock in 2005 has the heating carriers sharereported in Table 3, and the unchanged stock follows the trendfrom the same table until 2035. In all scenarios until year 2010also new and renovated buildings follow the same trend as thestock. � Base and Base+Conservation scenarios: the above assumptions
are valid until 2035.
� In the other scenarios a transition period is set from 2010 to
2020 for new and renovated buildings, in the same way asdescribed above. The initial state is the same as for the Basescenario; the final state is described below for each scenario. Inaddition, starting in 2010 also part of the stock is graduallyconverted to the new heating carriers share. It is assumed thatby 2035 half of the unchanged stock will have the heatingcarriers share described below.The transition period is meant to represent some inertia tochange in the energy system. This can be due either to thenecessity to set up an infrastructure, as for district heating orgas, or simply to the fact that new habits take time to maketheir way into society.
� Thermal and Thermal+Conservation scenario: buildings in the
final state will have a heating carriers share of 25% for directelectricity; the remaining 75% being covered by thermalcarriers.The choice of keeping some use of direct electricity is due tothe fact that not all locations can be covered by gas or districtheating networks, and not all locations are suitable for wooduse. Besides, electric heating can be combined with wood

ARTICLE IN PRESS

Table 6Scenario assumptions on Energy class and Heating carriers share.

Energy class Heating carriers share

Base, Th., HP Base, Th., HP + Conservation Base w/o Conservation Th., HP w/o Conservation

2005 2005 2010a 2020 2005 2005 2010 2020

2035 2010 2020 2035 2035 2010 2020 2035

Residential

Stock E-E* E-E* Trend Trend 1/2 trend

1/2-new

Renovation E* E* C-B B Trend Trend - New

New D D C-A A Trend Trend - New

Service

Stock E E Trend Trend 1/2 trend

1/2-new

Renovation E E C-B B Trend Trend - New

New E E C-A A Trend Trend - New

a In 2010 the EPBD will start giving its effect.


stoves when hydronic systems are not installed. Also, directelectric heating is a flexible and efficient solution for spacesthat need to be heated intermittently, because of its smallthermal inertia. Regarding the 75% from thermal carriers, noassumption is made on whether gas, wood or district heating isused. As the main goal is to study the substitution of electricitywith thermal carriers in general, it is not critical what theactual contribution of each single thermal carrier is. Concern-ing the efficiency, an average between gas, wood and districtheating efficiencies is taken. Oil is, however, completely phasedoff.
� Heat Pump and Heat Pump+Conservation scenario: these are
‘‘all-electric’’ scenarios and the buildings in the final state willhave a heating carriers share as follows. In the service sectorheat pumps will cover 75% of the heating need and directelectricity only 25%. In the residential sector heat pumps willcover 50% of the heating need while direct electricity coversthe other 50%.The use of direct electricity is meant to cover peak loads.Besides, some rooms may require higher comfort levels thatare difficult to achieve by heat pumps, i.e. bathrooms inNorway are typically equipped with electrical floor heating forcomfort reasons. In residential units the demand for hot wateris higher than in non-residential buildings, and heat pumps donot normally cover the domestic hot water demand.

All the assumptions on energy class and heating carriers shareare summarised in Table 6.

2 The reference year 2005 has the actual stock of year 2005 but energy class

and heating carriers share defined as the trend value of the observation period

1996–2005.

5. Results and discussion

An overall view is given in Fig. 4 where the past and theprojected delivered energy demand are represented. The figuresrefer to the entire building stock, i.e. both the residential andservice sectors together. All results refer to the middle level ofrenovation activity. The scenarios including energy conservationmeasures are those on the right hand side.

A summary of the model’s results for every scenario is given inTable 7. The data are presented for type of carrier, whetherelectricity or the sum of all thermal carriers, for sector, whetherresidential or service, and for building status, whether unchangedstock, renovation or new construction.

Fig. 5 compares the results for all scenarios as in 2035; thesituation in the reference year 2005 is also reported forcomparison.2 The total is divided into electricity and the sum ofthermal carries. The percentage reported inside the grey bars isthe percentage of electricity share. The dashed line represents theelectricity demand for the reference year 2005.

With respect to the reference year 2005 the Base scenariopresents an increase in total energy demand in 2035 of about8 TW h/y, divided in: +1 TW h/y in electricity demand and +7 TWh/y in thermal carriers.

In the Thermal scenario the total demand is the highest of all,about 15 TW h/y higher than 2005’s value, as the result of twoopposite trends: electricity demand �9 TW h/y and thermalcarriers demand +24 TW h/y. Altogether, the share of electricitygoes as far down as 53%, the lowest for all scenarios. Hence, thedependency on electricity is reduced both in relative and inabsolute terms with respect to the reference year 2005, but thisachievement has to be paid by a significant increase in the use ofthermal carriers.

The opposite situation is observed for the Heat Pump scenario.Here the total energy demand actually decreases by to �2 TW h/y,again as the combination of two diverging trends: +4 TW h/yelectricity and �6 TW h/y thermal carriers demand. The share ofelectricity is 85%, the highest for all scenarios. The electricitydependency has grown both in percentage and absolute valuefrom the reference year 2005, but the overall demand is thesmallest amongst the scenarios without conservation measures.

Hence, both the Thermal and the Heat Pump scenarios seem tohave both a positive and a negative outcome. On one hand it ispossible to reduce electricity dependency at the price of highertotal consumption; on the other hand it is possible to limit the totalconsumption at the price of increasing electricity dependency.

In the three scenarios with conservation measures thedifferences are smaller. In all of these scenarios the electricitydemand is lower than in the reference year 2005, with aconsistent reduction that varies between �8 and �16 TW h/y.The effect on thermal carriers varies between �6 and +12 TW h/yfrom the 2005 reference, but does never counterbalance thepositive achievements in electricity demand reduction. The

ARTICLE IN PRESS

scenario: Base

0

100

200

300

400

500

600

1975

Are

a [ M

illio

n m

2 ]A

rea

[ Mill

ion

m2 ]

Are

a [ M

illio

n m

2 ]

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

scenario: Base + Conservation

0

100

200

300

400

500

600

Are

a [ M

illio

n m

2 ]A

rea

[ Mill

ion

m2 ]

Are

a [ M

illio

n m

2 ]

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

scenario: Thermal

0

100

200

300

400

500

600

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

scenario: Thermal + Conservation

0

100

200

300

400

500

600

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

scenario: Heat Pump

0

100

200

300

400

500

600

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

scenario: Heat Pump + Conservation

0

100

200

300

400

500

600

0

10

20

30

40

50

60

70

80

90

100

Ene

rgy

[ TW

h/y

]

electricity thermal carriers free heat HP floor area

1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035

Fig. 4. Delivered energy results for the different scenarios. On the left side: (a) Base, (b) Thermal and (c) Heat Pump. On the right side the same scenarios plus energy

conservation measures: (d), (e) and (f), respectively.


scenarios with energy conservation show reduction in bothelectricity and total delivered energy. This means that the energyconservation approach is robust because it can guarantee, withina certain margin, the achievement of target results even underconditions of wide variation in the share of heating carriers.

On the other hand, it may be argued that the effect of the otherapproaches is dampened by the adoption of conservationmeasures, creating a contradiction de facto. For example – asanalysed in detail by Thyholt (2006) – on the supply side theprofitability of developing a district heating network would bereduced when the overall demand of a given area is reduced.Infrastructures for district heating or gas are more profitable in

areas with higher energy density. Therefore, large investments inthe expansion of such infrastructures would conflict withinvestments in energy efficiency.

This is described in Nassen and Holmberg (2005) as theproblem of ‘‘permanent lock-in’’, where the choice betweensupply and demand strategies always would favour substitutionof fuels at the expense of energy efficiency. It is opinion of theauthors that to overcome this risk, the building stock shouldrather be seen as an energy infrastructure itself, and energypolicies should aim at maximising its efficiency.

It may be argued that the total savings (5–15 TW h/y) arerelatively small because they lie in the range 7–21% compared

ARTICLE IN PRESS

Table 7Delivered energy for all the scenarios.

Delivered Energy by year 2035 (TW h/y)

Scenario Total Electricity Thermal carriers Residential Service Stock unchanged Renovation New

Reference year 2005 71.8 55.5 16.3 45.3 26.4

Base 79.9 56.5 23.4 47.8 32.1 41.4 19.9 18.5

Thermal 86.7 46.0 40.8 51.4 35.4 44.1 22.1 20.5

Heat Pump 69.7 59.1 10.6 41.7 28.0 37.4 16.7 15.6

Base+Cons. 63.1 45.7 17.4 38.7 24.4 41.4 12.1 9.5

Thermal+Cons. 67.2 39.1 28.1 41.0 26.2 44.1 13.0 10.1

Heat Pump+Cons. 56.9 47.2 9.7 34.7 22.2 37.4 10.8 8.7

Totals and grouped by different criteria.

Delivered Energy, per Scenario

77%

0

10

20

30

40

50

60

70

80

90

Referenceyear 2005

Base Thermal +Conservation

Heat Pump +Conservation

Ene

rgy

[ TW

h/y

]

Electricity Thermal carriers

71% 53% 85% 72% 58% 83%

Thermal Heat Pump Base +Conservation

Fig. 5. Total delivered energy by 2035 for each scenario, divided into electricity and sum of thermal carriers.

Table 8Total energy and electricity demand variation (2035 vs. 2005) and deviation due to

different levels of renovation activity.

Total (TW h/y) Electricity (TW h/y)

7Renovation 7Renovation

Middle Low High Middle Low High

Reference year 2005 – – – – – –

Base 8.1 +0.0 +0.0 1.0 +0.0 +0.0

Thermal 15.0 �0.1 +0.1 �9.5 +0.2 �0.2

Heat pump �2.1 +0.2 �0.2 3.6 +0.0 +0.0

Base+cons. �8.7 +2.6 �2.6 �9.8 +1.6 �1.6

Thermal+cons. �4.6 +2.9 �2.9 �16.4 +1.1 �1.1

Heat pump+cons. �14.9 +2.1 �2.1 �8.3 +1.7 �1.7


with the reference year 2005. However, it shall not be forgottenthat these savings take place while the building stock increases, asillustrated in Fig. 4 where the total energy curve is shown togetherwith the floor area curve. In the historical records the growth infloor area is closely accompanied by an equivalent increase inenergy demand. This behaviour is expected to continue unlessconservation measures are adopted. In that case a reverse trendcan be expected, and both electricity and total energy demand willdecrease, slowly but steadily, even though the stock keeps ongrowing. In 2035 the building stock in the model will be some 108million m2 more than in 2005, i.e. about 24% more. While thebuilding stock grows 24% larger, the energy demand is expected tobecome 7–21% smaller.

The results discussed so far are valid within the uncertaintiesassociated to the assumptions. As mentioned, estimates onrenovation activity are affected by high uncertainty because ofthe scarce record of data available. Three levels were defined forthis parameter, see Table 5, and their influence on the results isshown in Table 8. With respect to the middle level, the low andhigh levels would cause only negligible variations in the scenarioswithout conservation; while in those with conservation measuresthere would be a variation of 72–3 TW h/y in the total energy,and between 71–1.5 TW h/y in the electricity figures. Theseare significant variations because it means the deviation inpotential energy savings is in the range 715–30% for theBase and Heat Pump scenarios with Conservation, for both totalenergy and electricity demand. In the Thermal scenario with

Conservation – where the potential savings are the smallest –deviation in electricity savings is only 77% but deviation in totalenergy savings is as much as 764%.

The results for the residential sector can be comparedwith those from Myhre (2000); the numbers are summarised inTable 9. The initial conditions are somewhat different. Myhre’sestimate on stock consumption is higher than what is estimated inthis work in the starting year, when his stock is smaller. Thisdifference is merely due to different estimations; no actual(significant) energy efficiency improvement has occurred between

ARTICLE IN PRESS

Table 9Comparison with Myhre (2000).

Stocka (Million m2) Total energy

(TW h/y)

Electricity (TW h/y) Original stock

unchanged (%)

Starting year

Myhre (2000) 1998 211 49.0 35.0 100

This work 2005 220 45.3 34.3 100

Final year

Myhre (2000), reference 2030 291 60.0 47.0 50

This work, base 2035 270 47.8 34.8 66

Final year

Myhre (2000), high energy efficiency 2030 291 36.0 29.0 0

This work, Base+conservation 2035 270 38.7 28.6 66

a Heated floor area.


1998 and 2005. Furthermore, Myhre assumes the stock to increase38% in 32 years, while here it is assumed to increase 23% in 30years (residential stock only). Therefore, it is natural that Myhrewill estimate an even higher consumption when his stock growslarger, in the final year. That being stated, both total and electricitydemand are considerably higher in Myhre’s reference scenario. Itshall also be noticed that he makes the hypothesis that all newconstruction use only electricity for heating. In the best performingscenarios, however, the improvements are stronger in Myhre’swork, leading to absolute values that are similar to those of thiswork, despite his projected stock being larger. This is due toassumptions on the renovation activity. In Myhre’s work theportion of stock that remains unimproved after 32 years is assumedbeing 50% in the reference scenario, but in the high energyefficiency scenario the entire stock is converted to high standards ofenergy efficiency. This means that Myhre assumes different levelsof renovation activity in his scenarios. In this work, instead, theportion of unchanged stock is 66% in all scenarios. The authorsassume the renovation activity being constant in all scenarios;whether or not a building is to be renovated is something that goesbeyond considerations on energy efficiency. Renovation is simply asuitable moment for introducing energy efficient solutions.

The results of this work are not directly comparable with thoseof Johansson et al. (2006, 2007), who make scenarios for theSwedish building stock. However, it is worth comparing some ofthe assumptions. One of their scenarios assumes a 30% decrease inheat demand in the existing stock. The measures taken to achievesuch improvement are said to be economically competitivemeasures that include: replacement of windows, additionalinsulation, new ventilation systems with heat recovery andreplacing of heating system. These estimates on the intensitylevel appear absolutely reasonable. The problem is that the entirestock is assumed to undergo such changes in intensity level in just22 years, from 2003 to 2025 (the simulated period). Because theassumed demolition activity is negligible (0.03% annually), thatwould imply that 4.5% of the stock have to be renovated everyyear. Equivalently, adopting the assumption of this work that partof the stock remains unchanged while part is renovated to a highstandard of energy efficiency (class B) leads to the conclusion that48% of the stock should be renovated in 22 years in order to obtainthe same overall reduction in energy demand. Annualizing, thatbecomes a renovation rate of 2.1% per year, which is seven timeshigher than the declared construction rate of 0.3% per year. Thismay well be the case in the Swedish building stock, but it showsthat it is important to state explicitly both intensity and activitylevels, and distinguish clearly what is part of the commonassumptions and what is scenario specific.

6. Conclusions

A model has been developed based on the notions ofarchetypes and activity flows. In this model the activities ofconstruction, demolition and renovation are explicitly defined andare independent from the scenario analysis. The situation isdifferent in other works, where improvements in the existingstock are applied, either explicitly or implicitly, to differentportions of the stock in different scenarios. It is shown that thishas major consequences on the results, and therefore assumptionson activity and intensity levels should always be kept clearlyseparated and their effect analysed separately.

The evaluation of the three hypothetical approaches outlinedthe strength of adopting conservation measures against theconflicting effects of the other approaches based on thermalcarriers and heat pumps. Adopting conservation measures on alarge scale does allow reducing both electricity and total energydemand from present day levels while the building stock keepsgrowing.

Further work would be necessary to consider aspects like:climatic zones, type of area (urban/rural, high/low energydensity), disaggregation of building sectors into categories, agegroups and type of heating systems available. Extending theboundaries of the model is also necessary for allowing estimationof CO2 emissions. As an alternative, this analysis of the energydemand side could be used to provide input to models for theanalysis of the supply side. Supply side models usually assumeenergy demand as an exogenous variable; estimations made onthe basis of demand side models like the one presented herewould provide better guesses than a simple trend analysis of pastvalues.

Acknowledgments

This work was developed in the context of the TRANSES project(Transition to Sustainable Energy Services in Northern Europe), incollaboration between the Norwegian University of Science andTechnology (NTNU), SINTEF, Chalmers University of Technology–Goteborg, and the Massachusetts Institute of Technology (MIT).

Appendix A

See Appendix Tables A1 and A2.

ARTICLE IN PRESS

Table A1Intensity values for several types of building.

GAB code Type of

building

Rs,

delivered

energy

(kW h/m2y)

Rr, net energy (kWh/m2y)a Share of

floor area in

2005 (%)

Total Heating Ventilation Hot waterb Fans and

pumpsb

Lighting Equipment Room

cooling

Ventilation

cooling

11x, 12x, 13x Small house 229 129 45 6 30 (37) 8 (1) 17 23 0 0 80

14x, 15x Apartments

block

218 118 31 7 30 (40) 10 (1) 17 23 0 0 12

16x Holiday house 47 47 8

31x, 39x Office 233 165 34 21 5 22 25 34 0 24 23

32x, 33x Shop 390 237 44 34 10 42 56 4 0 47 16

4xx Transport &

comm.

246 5

5xx Hotel &

restaurant

296 239 61 29 30 35 47 6 0 31 7

61x School 194 137 40 27 10 25 22 13 0 0 13

62x, 63x University 179 34 24 5 27 25 34 0 30 2

612 Kinder garden 224 152 67 26 10 23 21 5 0 0 0

65x Sport 279 185 48 40 50 23 21 3 0 0 4

64x, 66x, 67x,

69x

Culture 188 178 66 26 10 24 23 3 0 26 8

71x Hospital 390 327 57 42 30 54 47 47 0 50 2

72x, 73x, 79x Nursery 279 234 48 38 30 48 47 23 0 0 5

8xx Prison etc. 249 2

9xx Other 13

a End use values are indicative estimations.b Values referring to the existing stock (Rs, delivered energy), when they are known, are reported in parentheses.

Table A2Efficiency or COP associated with energy carriers.

Electric needs Cooling needs Heating needs

Electricity Electricity Oil Gas Wood District heating Electricity Heat from HP

Residential 1.00 2.40 0.77 0.81–0.86a 0.4–0.6a 0.88 1.00 2.16

Service 1.00 2.40 0.77 0.86 0.70 0.88 1.00 2.32

a Left value in existing stock, right value in new and renovated.


Appendix B

See Appendix Tables B1–B4.

Table B1Archetype for residential sector, stock reference Rs, observed heating carriers share.

Class E, Rs


Energy need kW h/

m2y

Share (%) Carrier Heating (%) kW h/

m2y

Efficiency

or COP

Carrier kW h/

m2y

Energy carrier kW h/

m2y

Share (%)

190.0 Carriers share 212.8

Electric 41.0 21.6 1.00 Electricity 41.0 Electricity 162.0 76.1

Cooling 0.0 0.0 2.40 Electricity 0.0

Heating 149.0 78.4 Electricity direct 80.7 120.2 1.00 Electricity direct 120.2

District heating 0.9 1.4 0.88 District heating 1.6 District heating 1.6 0.7

Wood 8.9 13.2 0.40 Wood 33.0 Wood 33.0 15.5

Gas 0.3 0.5 0.81 Gas 0.6 Gas 0.6 0.3

Oil 8.1 12.1 0.77 Oil 15.7 Oil 15.7 7.4

Heat from HP 1.1 1.7 2.16 Ele. to drive HP 0.8

Free heat HP 0.9 Free heat HP 0.9 –

ARTICLE IN PRESS

Table B2Archetype for residential sector, regulation reference Rr, observed heating carriers share.

Class C, Rr


Energy need kW h/m2y Share (%) Carrier Heating (%) kW h/m2y Efficiency or COP Carrier kW h/m2y Energy carrier kW h/m2y Share (%)






Wood 8.9 6.5 0.40 Wood 16.2 Wood 16.2 12.2

Gas 0.3 0.2 0.81 Gas 0.3 Gas 0.3 0.2

Oil 8.1 5.9 0.77 Oil 7.7 Oil 7.7 5.8



Table B3Archetype for service sector, stock reference Rs, observed heating carriers share.

Class E, Rs








Wood 0.1 0.1 0.70 Wood 0.2 Wood 0.2 0.1

Gas 0.9 1.5 0.86 Gas 1.7 Gas 1.7 0.7

Oil 15.4 25.5 0.77 Oil 33.1 Oil 33.1 12.6



Table B4Archetype for service sector, regulation reference Rr, observed heating carriers share.

Class C, Rr








Wood 0.1 0.1 0.70 Wood 0.1 Wood 0.1 0.1

Gas 0.9 0.8 0.86 Gas 0.9 Gas 0.9 0.5

Oil 15.4 13.2 0.77 Oil 17.2 Oil 17.2 9.5




References

Bøeng, A.C., 2005. Energibruk i husholdninger 1930–2004 og forbruk etter

husholdningstype, Statistics Norway, Oslo.

DGET, 2004. European Union Energy and Transport in Figures—part 2: Energy,

Directorate-General for Energy and Transport.

ECBCS, 2004. Stock aggregation, in Annex 31—Energy related environmental

impact of buildings, IEA-ECBCS (International Energy Agency–Energy Con-

servation in Buildings and Community Systems).

Eggen, G., 2005. NVE-Varmepumpens bidrag til redusert energibruk i Norge,

Utredning gjennomført av COWI for NVE, datert 12.2005.

EN 15217, 2007. Energy performance of buildings—Methods for expressing energy

performance and for energy certification of buildings, CEN.

Enova, 2006. series of Bygningsnettverkets energistatistikk: 2002, 2003, 2004,

2005, 2006, Enova, Trondheim.

EPBD, 2002. Directive 2002/91/EC of the European Parliament and of the Council of

16 December 2002 on the Energy Performance of Buildings, Official Journal of

the European Communities, 4.1.2003, pp. L1/65–L1/71.

ARTICLE IN PRESS


Grorud, C., Rasmussen, I., Strøm, S., 2007. Fremskrivning av varmepumpens bidragtil den Norske energiforsyningen, Utredning for NVE av Vista Analyse AS,datert 14.2.2007.

Johansson, P., Nylander, A., Johnsson, F., 2006. Electricity dependency and CO2

emissions from heating in the Swedish building sector—current trends inconflict with governmental policy? Energy Policy 34, 3049–3064.

Johansson, P., Nylander, A., Johnsson, F., 2007. Primary energy use for heating in theSwedish building sector—current trends and proposed target. Energy Policy 35,1386–1404.

Myhre, L., 2000. Towards Sustainability in the Residential Sector, Byggforsk, NBInote 41.

NVE, 2008. Statistikk og analyser, Import og eksport 1990–2006, Norges vassdrags-og energidirektorat, /www.nve.noS.

Nassen, J., Holmberg, J., 2005. Energy efficiency—a forgotten goal in the Swedishbuilding sector? Energy Policy 33, 1037–1051.

Pettersen, T.D., Myhre, L., Wigenstad, T., Dokka, T.H., 2005. Energimerking avboliger, Byggforsk Report for NVE, O 20461.

Rødseth, A., Barlindhaug, R., Østervold, J., 1997. A simulation model of the Norwegianhousing market, draft, Department of Economics, University of Oslo, Oslo.

Sartori, I., 2006. ePlan-Data on Activity, SINTEF Byggforsk, Project memo, dated29.9.2006.

Sartori, I., Wachenfeldt, B.J., 2007. ePlan 2006 Final Report, SINTEF Byggforsk,Rapport SBF BY A07009.

Sartori, I., Bergsdal, H., Muller, D.B., Brattebø, H., 2008. Toward modelling ofconstruction, renovation and demolition activities: Norway’s dwelling stock1900–2100. Building Research and Information 36 (5), 412–425.

Statistics Norway, 2001. Series of Population and Housing Census (folke- ogboligtellingen): 1980, 1990, 2001, Statistics Norway, Oslo.

Statistics Norway, 2007a. Natural resources and the environment, Resources,Energy, Statistics Norway, available at /www.ssb.no/english/subjects/01/03/10/S.

Statistics Norway, 2007b. Industrial activities, Construction, Statistics Norway,available at /www.ssb.no/english/subjects/10/09/S.

Statistics Norway, 2007c. Population, Statistics Norway, available at /www.ssb.no/english/subjects/02/S.

Steen, P., Johansson, T.B., Fredriksson, R., Bergren, E., 1981. Energi—till vad och hormycket? Stockholm.

TEK, 2007. Veiledning til TEKnisk forskrift til plan- og bugningsloven 1997–4.utgave mars 2007, Statens Bygningstekniske Etat, Oslo.

Thyholt, M., 2006. Varmeforsyning til lavenergiboliger i omrader med fjernvarme-konsesjon: analyser av CO2-utslipp og forsyningssikkerhet for elektrisitet,Doctoral Thesis, NTNU, 2006:162 (ISBN 82-471-8092-8).

Thyholt, M., Dokka, T.H., Uvsløkk, S., Gustavsen, A., 2007. Nye energikrav i tekniskeforskrifter, Sintef Byggforsk, SBF BY A07002.

Wachenfeldt, B.J., 2004. A spreadsheet tool for scenario analysis of the energyconsumption in buildings and resulting emissions, SINTEF Civil and Environ-mental Engineering, TRANSES Project Memo, dated 18.05.2004.

Wachenfeldt, B.J., Sartori, I., 2007. Varme og energietterspørsel i privatehusholdninger og tjenesteytende sektor frem mot 2035, SINTEF Building andInfrastructure, Project Memo, Project 3B007200, dated 14.9.2007.

Wigenstad, T., Dokka, T.H., Pettersen, T.D., Myhre, L., 2005. Energimerking avnæringsbygg, Sintef Teknologi og Samfunn Report for NVE, STF50 F05117.

http://www.nve.no

http://www.ssb.no/english/subjects/01/03/10/

http://www.ssb.no/english/subjects/01/03/10/

http://www.ssb.no/english/subjects/10/09/

http://www.ssb.no/english/subjects/02/

http://www.ssb.no/english/subjects/02/

energy demand in the norwegian building stock: scenarios ... · energy demand in the norwegian...

Documents