Weekly greenhouse gas emissions of municipalities: Methods and comparisons

Energy Policy 39 (2011) 4755–4765

Contents lists available at ScienceDirect

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

Weekly greenhouse gas emissions of municipalities: Methods and comparisons S. Monni a,n, S. Syri b a b

Benviroc Ltd, Lekkerikuja 1 B 21, FI-02230 Espoo, Finland Aalto University, School of Engineering, Department of Energy Technology, P.O. Box 14100, FI-00076 Aalto, Finland

a r t i c l e i n f o

abstract

Article history: Received 23 February 2011 Accepted 23 June 2011 Available online 16 July 2011

Local authorities need timely information on their greenhouse gas (GHG) emissions and their causes, comparison with other municipalities and tools for dissemination of information to the citizens. This paper presents a weekly GHG emission calculation system, CO2-report, which provides such data for citizens and local decision-makers in a timely manner, in contrast to the official emissions statistics, which are available on an annual basis 1–2 years afterwards. In this paper, we present the methodology and three main outputs of CO2-report: (1) weekly GHG emissions; (2) advance annual emissions; and (3) final annual emissions for 2009 with comparison of 64 municipalities in Finland. We explain the reasons for the large variability of annual emissions, from 5 to 13 t CO2-eq/capita, discuss the accuracy of advance and final emission estimates at local level, and show the weekly variability of emissions for three example municipalities with different emission profiles. & 2011 Elsevier Ltd. All rights reserved.

Keywords: Greenhouse gas emissions Local emissions Weekly emissions

1. Introduction As the concern on the impacts of climate change has spread from the scientific community to the general public, residents of cities and municipalities have become interested in emissions caused by their daily activities. Many tools have been developed in recent years to describe these impacts. Efficient and timely communication of the results of local GHG inventories to local decision makers and citizens is important, as the ultimate goal of emission inventories is to support the emission reduction actions. Information on local GHG emissions and their causes can be provided to citizens by developing tools for estimating real time emissions and utilizing the possibilities of present ICT to provide information for residents and decision-makers on-line. One emission information tool that has rapidly gained popularity is the calculation of the so-called carbon footprint. Carbon footprint is based on the greenhouse gas emissions of a product or service over its entire life cycle, being an illustrative tool when different products or services or different stages in the life cycle are compared. However, carbon footprints of cities are dataintensive, and therefore city greenhouse gas (GHG) emissions have typically been quantified based on an inventory approach, which includes annual emissions occurring within the city, and the methods of treating extra-boundary activities such as

n

Corresponding author. Tel.: þ358 40 543 1476. E-mail addresses: Suvi.monni@benviroc.fi (S. Monni), sanna.syri@aalto.fi (S. Syri). 0301-4215/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.enpol.2011.06.051

material use and transportation may vary. Imports and exports of electricity and heat across city borders are also taken into account, but the methodologies for accounting such imports and exports vary (e.g. Sovacool and Brown, 2010; Kennedy et al., 2010; Ramaswami et al., 2008). The EU energy and climate package has set goals to reduce GHG emissions by 20% by 2020, with an option to increase the reduction target to 30% if a comprehensive international agreement is reached (Council of the European Union, 2007). In addition, the EU has set targets for renewable energy use and increased energy efficiency. For Finland, the EU obligation is to reduce the GHG emissions from the sectors outside the EU emissions trading system by 16% by the year 2020 from the 2005 level. In order to achieve the EU target, the Finnish Government prepared National long-term Climate and Energy Strategy in 2008 (Ministry of Employment and the Economy, 2008). According to the strategy, Finland’s emissions in 2020 would be about 20% above the 1990 levels if new climate policy measures were not carried out. Therefore, achieving the targets requires a wide array of measures in e.g. energy conservation, energy efficiency and renewable energy use in energy, industry and transport sectors. Finland’s obligation to increase the share of renewable energy is from the present about 28% to 38% of final energy use by 2020. Also this target is challenging, and achieving it requires that final energy consumption is turned to a declining trend and fossil fuels are replaced by renewable energy. Finland’s ambitious emission targets cannot be achieved without local-level commitment to mitigating climate change. Municipalities are often in a position to make decisions that affect local

4756

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

emissions in the short, medium and long term. In particular, municipalities are responsible for land-use and transport planning, building permits, waste management and providing local public transportation services. Some of the municipalities are major local energy suppliers or owners of energy supply companies, even though this role has changed in many cities during the last ten years due to the privatization and liberalization processes of the electricity market in the Nordic countries. The municipalities are also close to the people, and can significantly influence how residents consume, for example by information measures. The municipalities can also promote renewable energy by removing barriers through, for example, land-use planning and building permits which favor renewable energy. In this article we present the development of a weekly on-line greenhouse gas emission monitoring system for municipalities and cities in Finland, called the CO2-report. CO2-report provides timely information of local emissions and their sources for citizens and decision-makers in contrast to the official emissions statistics, which are available on an annual basis 1–2 years afterwards. In Finland, the official annual emissions statistics are compiled by Statistics Finland, and they are published usually in December the following year. In the present phase of the development of CO2-report, the information is made available for residents via web pages, which are updated always on Mondays–Tuesdays of the following week. In addition, there are wide further development opportunities, for example using mobile applications. In addition to weekly emission estimates, the system provides an annual advance GHG estimate by municipality, and when the final annual statistics are available, the system calculates the final GHG emission figures for the municipalities. Currently the emissions of 64 municipalities are calculated in the system using the same methodology, covering 58% of population in Finland. This provides a unique dataset of GHG emissions of Finnish municipalities, facilitating understanding of the most important sources of emissions and their mitigation options. Section 2 describes the methods, data sources and processing of data used in the modeling. Section 3 presents the calculation results, discusses the reasons for differences between municipalities and provides an analysis of uncertainties. Annual emission estimates are presented for all municipalities included and weekly estimates are shown for three examples, Helsinki, Parikkala and Kuhmoinen, which have very differing emission profiles, Helsinki being the capital of Finland, Parikkala being a typical rural community with significant agricultural activity and Kuhmoinen showing the impacts of a large amount of summer cottages in the community.

2. Methods The calculation model ‘‘CO2-report’’ calculates weekly emissions for more than 60 municipalities, who have joined to the system, out of about 340 municipalities in total in Finland. The basis of weekly GHG estimates for the year n (for example 2009) is the annual GHG estimate calculated for the latest year for which official statistics are available (year n  1 or n  2). Based on the weekly indicators described below, the weekly GHG emissions are calculated for each week of year n, which at the end of the year constitute the advance emission estimates for the entire year. When the official statistics are updated for year n (in year n þ2), the final emission estimates for year n are calculated, and the emissions of all over 60 municipalities are compared. The methodology corresponds at municipality-level Finland’s official methodology of reporting national emissions to the UN FCCC (Statistics Finland, 2010a). The model comprises the three

most important greenhouse gases: carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). CH4 and N2O emissions are calculated with the global warming potential factors 21 and 310, respectively, as carbon dioxide equivalent (CO2-eq). Weekly greenhouse gas emissions are calculated for seven sectors, which are of particular interest for consumers and decision makers at the local level:

      

District heating Building-specific heating Electric heating of buildings Other electricity consumption (excluding industry) Road transport Agriculture Waste.

Emissions of industrial activities, other transport modes and emissions and removals from land use may be included in annual reports provided to municipal authorities, but are not included in the weekly statistics. The CO2-report calculates the emissions of activities, which occur inside the municipality’s geographical boundaries, such as consumption of transport and heating fuels, cultivation of land and animal husbandry. However, the emissions related to energy and waste are allocated to municipalities based on consumption of electricity and heat, and generation of waste and wastewater. Main data sources, data processing steps and key assumptions by category are summarized in Table 1. 2.1. District heating Emissions of district heating describe the emissions of production of district heat which is consumed in the municipality. The fuels consumed in district heating are based on the statistics compiled by the Finnish Energy Industries (2010a) and on direct questionnaires for the district heat providers. Emissions from combined heat and power (CHP) production are allocated to electricity and district heat based on the emissions of an alternative (separate) production of heat and electricity (European Union/Covenant of Mayors, 2010). This is a commonly used method in the Nordic countries out of several possible approaches (e.g. Graus and Worrell, 2011). CO2 emission factors of fuel use are those compiled by Statistics Finland, and the CH4 and N2O emission factors are from the Kasvener model developed at the Finnish Environment ¨ a, ¨ 2007). The fuel-specific emission factors of the Institute (Petaj Kasvener model are average emission factors by boiler type (e.g. district heat boiler) instead of technology (e.g. grate fired boiler), as in the official GHG inventory. The weekly variation of emissions from district heating is based on the variation of heating requirement, which is measured by heating degree days. The Finnish Meteorological Institute calculates the weekly and monthly heating degree days at 20 measurement stations across Finland. The heating degree day number is calculated as the daily average difference between the outdoor and indoor temperatures, the indoor temperature being assumed as þ17 1C (the so-called S17 heating degree day). In the CO2-report we use the measurement station closest to the municipality in concern, and calculate the weekly heating-degree days based on the average heating demand in the municipality compared to the measurement station location. District heating suppliers of several municipalities provide information on the monthly or annual fuel mix used for district heat production. For the rest of the municipalities, the latest fuel mix in district heating statistics is used as a basis in the weekly calculation of emissions until final statistics are available.

Table 1 Data sources, data processing steps and key assumptions used to compile the final GHG estimates of municipalities. ‘‘Data processing’’ refers to steps taken to modify the data so that the emission factors and parameters used in the Finnish GHG inventory can be applied. Data

Source

Data processing

Key assumptions

District heating

Fuel consumption by district heat supplier

Finnish Energy Industries (2010a); questionnaires to district heat producers

Purchases and sales of heat across borders of municipalities taken into account; fuel use in CHPn plants divided between electricity and heat

Building-specific heating (oil and gas)

Floor space by building type and heating system (m2)

Statistics Finland (2010b)

Building-specific heating (wood) Electricity consumption

Amount of wood combusted (m3)

Finnish Forest Research Institute

Electricity consumption by municipality in industry and other uses

Finnish Energy Industries (2010c)

Road transportation

Emissions by municipality

Enteric fermentation

Number of animals

Manure management, manure in pasture and use of manure as fertilizer Nitrogen fertilization and liming Crop residues and N fixing crops Landfilling of waste (landfills in operation)

Number of animals

VTT Technical Research Centre of Finland (2010) Information Centre of the Ministry of Agriculture and Forestry; Finnish Trotting and Breeding Association; Reindeer Herders’ Association Information Centre of the Ministry of Agriculture and Forestry; Finnish Trotting and Breeding Association; Reindeer Herders’ Association Information Centre of the Ministry of Agriculture and Forestry Information Centre of the Ministry of Agriculture and Forestry (a) VAHTI database; Kuittinen et al. (2010): questionnaires on start years of landfill operations (b) Questionnaires to landfill operators

Floor space by building type multiplied with average energy consumption for space heating and production of hot water (calculated based on different data sources, see Section 2.2) taking into account heating degree days of the municipality Data in m3 wood converted into energy units by type of wood fuel Electricity consumption in other sectors than industry divided to ‘‘electric heating’’ and ‘‘other electricity consumption’’ based on modeling of heating demand of buildings (see Section 2.3) None

Fuels are combusted in ‘‘a typical district heat boiler’’ or ‘‘a typical CHP plant’’, i.e. data are not divided by technology (but by fuel). This assumption has only impact on CH4 and N2O emissions, not on CO2 Energy needed for heating (kWh/m2) depends only on building type and location (i.e. age, construction material, design, use, etc. not taken into consideration)

Landfilling of waste (closed landfills)

Waste disposed to landfills, start and closing year of the landfill, landfill gas recovery Biochemical oxygen demand (BOD7) load to municipal WWTPs

Municipal wastewater treatment plants (WWTPs), CH4 Industrial WWTPs, CH4 Municipal and industrial WWTPs, N2O Wastewater from rural areas Composting n

Cultivated area Cultivated area by crop type (13 types)

(a) Waste disposed to landfills, amount of LFGnn collected, start year of the landfill operation (b) Emissions from landfills

Questionnaires to environmental administrators of the municipality

None

None

None None

(a) Emissions calculated with a first order decay model and divided to municipalities served by the landfill based on population (b) Emissions divided to municipalities served by the landfill based on population

Manure management systems are those used in Finland in average; all manure excreted in the municipality is applied to the agricultural fields in the same municipality Application of N fertilizers and lime (kg/ha) is Finnish average Share of crop residues left to soils is Finnish average (a) Historical development of waste amount by waste type is Finnish average for years for which no data are available (b) See (a) as most of landfill operators use the same model and assumptions

VAHTI database

If only total amount of waste is known, divided between main waste types; emissions calculated with a first order decay model None

Finnish averages on waste composition and historical development used if no better data are available None

Chemical oxygen demand (COD) of industrial wastewater into waterways N load to waterways

VAHTI database

COD to waterways converted to incoming COD

VAHTI database

None

Average treatment efficiency used to convert COD to waterways to incoming COD None

Population in rural areas

Statistics Finland

None

Amount of waste composted by type

VAHTI database

Different types of sludge converted to dry matter

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

Category

Population in rural areas represents the population not connected to municipal wastewater treatment system None

CHP¼ combined heat and power production. LFG¼ landfill gas.

nn

4757

4758

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

2.2. Building-specific heating The emissions from building-specific heating are calculated for all buildings within the municipalities, which are heated by fuel oil or natural gas. The energy needed to produce hot water is modeled based on information on the typical water use in different buildings available from Motiva Ltd. (which is the state-owned company specialized in informing about energy efficiency and energy saving in Finland) (Motiva, 2010). The building types included are for example residential, office and educational buildings. The fuel use for space heating of buildings is calculated by combining statistics on fuels used in separate heating of buildings in whole Finland (Statistics Finland, 2009) with municipality-level information on heating degree days, building stock (floor space) and building types (Statistics Finland, 2010b). The emissions from use of wood for space heating are calculated based on municipality-level statistics of Finnish Forest Research Institute, which carries out periodic surveys. 2.3. Electricity consumption Emissions from electric heating of buildings are calculated by combining national-level statistics on energy demand of electric heating of buildings (Statistics Finland, 2009) with municipalitylevel information on heating degree days, building stock and building types. The emission factor is the monthly average emission factor of electricity consumption for the whole of Finland, calculated based on the statistics of the Finnish Energy Industries (2010b) and Statistics Finland (2011). The emissions from CHP plans are divided between electricity and heat based on fuel required for alternative production (European Union/Covenant of Mayors, 2010). The data of Finnish Energy Industries (2010b) on fuels used for electricity production in Finland are based on fuel consumption of each electricity plant each month. However, the plant-specific fuel mix is preliminary data until June in the following year, when the plants report their final fuel consumption figures. The annual emission factor for electricity varies due to fuels used for electricity generation, and the balance of electricity imports and exports. In 2009, for instance, there was more coal consumption for electricity production than in 2008, and thus the emission coefficient for electricity consumption was higher in 2009 (196 g CO2/kWh) than in 2008 (173 g CO2/kWh). The term other electricity consumption comprises all other electricity consumption than heating in the following sectors: housing, services, agriculture and construction. The municipalitylevel electricity consumption modeled is based on the annual statistics of the Finnish Energy Industries (2010c). The weekly variability of electricity consumption is based on the statistics of Finnish Energy Industries (2010b) on total weekly electricity consumption in Finland. The entire variability is addressed to other electricity consumption, as the electricity consumption of industry is relatively stable throughout the year. 2.4. Road transportation Emissions from road transport calculated in the CO2-report cover all road transport taking place within the area of the municipality, thus both local traffic and through-traffic are included. Therefore, for instance, the municipalities through which main national highways pass have a high level of transport emissions in comparison to other municipalities. The official annual emissions of road transport by municipality and by vehicle type calculated by the LIISA model (VTT Technical Research Centre of Finland, 2010) are used as the starting point in CO2report. The week-level information on traffic volumes is based on the on-line monitoring of traffic volumes of the Finnish Transport

Agency (2010). The Finnish Transport Agency has about 350 transport monitoring devices across Finland. They are placed at all main roads and at main streets of major cities in Finland, and they provide hourly data on traffic volumes. The CO2-report calculates weekly emissions by assuming that the transport emission factors remain the same as in the previous year, until the final statistics are available. In reality, the CO2 emission factor decreases when the share of biofuels in transportation increases (when biofuel combustion is calculated to cause zero CO2 emissions). Furthermore, the switch from gasoline to diesel decreases emissions per km, and also the change in vehicle fleet has an impact on the average emission factor. According to the VTT Technical Research Centre of Finland (2010) the average CO2 emission factor decreased from 220 g CO2/km in 2008 to 207 g CO2/km in 2009. 2.5. Agriculture The emissions from the agriculture sector include emissions from agricultural activities within the municipality. The emission sources included are enteric fermentation, manure management, manure in pasture, use of synthetic fertilizers, liming, use of manure as fertilizer, N fixing crops, crop residues and indirect N2O from leaching and run-off. The activity data are based on the municipality statistics of Information Centre of the Ministry of Agriculture and Forestry (Tike) on animal numbers of cattle (5 types), sheep, goat, swine, poultry (5 types), total cultivated area and cultivation areas for 13 most important crop types. The animal number statistics of Tike are supplemented by the statistics of Finnish Trotting and Breeding Association (number of horses and ponies) and Reindeer Herders’ Association (number of reindeer). The use of nitrogen fertilizers in each municipality is calculated based on average fertilizer use in Finland and cultivated area. The emission factors and other calculation parameters applied are those of the Finnish GHG inventory (Statistics Finland, 2010a). The weekly variation of emissions is modeled based on the occurrence of agricultural activities throughout the year. The pasture season (about 120 days between May and August) implies higher emissions from enteric fermentation (higher energy demand of animals) and manure in pasture, and lower emissions from manure management systems. N2O emissions from manure application to soils occur relatively soon after spread of manure, and therefore the emissions of manure application are assumed to occur during the period when manure application to soils is allowed by law. The other agricultural emissions are assumed to occur at a constant rate across the weeks of the year. 2.6. Waste management Emissions from waste management include emissions from landfilling, composting and wastewater treatment. Most of the waste management companies calculate the CH4 emissions of their landfills using a first order decay (FOD) model developed by Finnish Environment Institute. The model includes the same parameters as the Finnish GHG inventory. In this case, the model results are used in CO2-report. In case such calculation is not available, the emissions are calculated using the Finnish Environment Institute’s FOD model based on the amount of waste disposed in landfills and data on landfill gas recovery (Kuittinen et al., 2010). Most of the landfills serve several municipalities. The landfill emissions are allocated to the municipalities based on their share of the population served by the landfill, as average waste generation per capita does not vary significantly between municipalities within a region. Emissions from closed landfill sites are also calculated using the FOD model, and based on information on waste amounts requested from the municipalities. However, the

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

study. The lowest emissions are in Raahe (Fig. 1), a city with 23 000 inhabitants. District heating in Raahe is produced mainly (75%) using industrial waste heat, which is considered emission free. There is no significant through traffic, and the emissions of local transportation are moderate. The importance of agriculture is also minor. However, there is heavy industry in Raahe, and having these emissions included, Raahe would not be the least emitting city. The highest emissions per capita occur in Hartola, which is a municipality of 3400 inhabitants. The main south-north highway in Finland, E75, goes through Hartola, causing the majority of the emissions in the municipality. Emissions of road transportation cover half of total emissions in Hartola, of which one third is

availability and quality of the data varies. The data on wastewater treatment and large-scale composting by municipality is obtained from the environmental administration’s VAHTI compliance data system. The emissions are calculated using the emission factors of the Finnish GHG inventory (Statistics Finland, 2010a).

3. Results 3.1. Annual emissions in 2009 The annual GHG emissions per capita in 2009 vary from 4.8 kg CO2-eq to 13.3 t CO2-eq in 64 municipalities included in the Waste

Agriculture

Road Transportation

4759

Other electricity consumption

Electric heating

Building-specific heating

8

10

District heating

Hartola Kiuruvesi Sysmä Hämeenkoski Ikaalinen Kuhmoinen Parikkala Jokioinen Eurajoki Hankasalmi Hämeenkyrö Loviisa Mäntsälä Kärkölä Padasjoki Orimattila Mynämäki Janakkala Salo Ylivieska Kuusamo Heinola Sipoo Hamina Nastola Hausjärvi Kemiönsaari Uusikaupunki Hollola Mikkeli Raisio Kuopio Kangasala Vihti Lohja Lappeenranta Masku Lahti Riihimäki Tuusula Äänekoski Kaarina Jyväskylä Hyvinkää Turku Rusko Kotka Oulu Pirkkala Vantaa Ylöjärvi Nurmijärvi Rauma Kauniainen Imatra Joensuu Kerava Järvenpää Kirkkonummi Pornainen Tampere Espoo Helsinki Raahe 0

2

4

6

12

t CO2-eq/cap Fig. 1. Final annual emission estimates in 2009 (t CO2-eq/capita) by sector in 64 municipalities of the CO2-report.

14

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

14 12

t CO2-eq/capita

10 8 6 4 2 0 100000

0

200000

300000

400000

500000

600000

population Fig. 2. GHG emissions per capita compared with population in 64 municipalities in Finland. The dashed vertical line shows the limit of 70 000 inhabitants.

120000

Agriculture Waste Distric heating Building-specific heating Electric heating Other electricity Road transportation

100000

80000 t CO2-eq

caused by trucks (VTT Technical Research Centre of Finland, 2010). In the 64 municipalities included in the study, the annual emissions of heating (district heating and building-specific heating) are from 0.6 to 3.5 t CO2-eq/capita. The emissions are highest in cities, in which coal (e.g. Turku, Raisio, Lahti) or peat (e.g. Kuopio, Ylivieska) is an important fuel for district heating production. In addition, the floor-space of buildings per capita is often higher in cities than in smaller municipalities, as the cities provide public and private services (such as hospitals, education, shopping centers) also for residents of surrounding municipalities. The emissions of electricity consumption, including both electric heating and other electricity consumption (excluding industry) vary between 1.2 and 2.8 t CO2-eq/cap. As the national emission factor is used for all the municipalities at monthly level, the differences of the emissions are due to electricity consumption and the share of electric heating in total electricity consumption. The emissions of road transportation vary from 0.9 to 6.7 t CO2-eq/capita, and are generally smaller in cities and larger in small municipalities with through traffic. The emissions of agriculture are between 0.01 and 7.6 t CO2-eq/capita, being negligible in cities, such as Helsinki, Espoo, Kerava and Kauniainen. On the contrary, in Kiuruvesi (population 9300), the most important beef producing and the second important milk producing municipality in Finland, emissions from agriculture cover 62% of total emissions. The emissions of waste management vary between 0.1 and 0.8 t CO2-eq/capita. The emissions are highest in municipalities with large landfills of pulp and paper industry, and smallest in cities in which the municipal landfills are equipped with efficient landfill gas recovery systems. The estimates of waste management emissions are not entirely comparable between different municipalities. Methane is formed also in closed landfill sites, and availability of information on these sites varies between municipalities. There are only few municipalities in Finland using waste ¨ since 2008). As there incineration (e.g. Hyvinka¨ a¨ and Riihimaki is no waste incineration without energy recovery in Finland, the emissions of waste incineration are allocated to electricity and district heat produced by waste combustion facilities, following the methodology of Finnish GHG Inventory (Statistics Finland, 2010a). Despite that, the waste management emissions of ¨ in 2009 are not smaller than in other Hyvinka¨ a¨ and Riihimaki municipalities, as the waste landfilled in earlier years still produces methane. When the emissions of closed landfill sites gradually decrease, the waste management emissions of municipalities with waste incineration will be smaller than those of municipalities with landfilling. In our sample of 64 municipalities, which cover population of 3.1 million (58% of population of Finland) the emissions per capita are generally larger in small municipalities than in cities (Fig. 2). However, the variation of emissions in small municipalities is large. In municipalities with less than 70 000 inhabitants, the emissions vary between 4.8 and 13.3 t CO2-eq/capita, whereas in cities with 470 000 inhabitants, the emissions vary between 4.9 and 6.9 t CO2-eq/capita.

60000

40000

20000

0 1

5

9

13

17

21

25

29

33

37

41

45

49

Fig. 3. Weekly GHG emissions in Helsinki in 2009.

1600

Agriculture Distric heating Electric heating Road transportation

1400

Waste Building-specific heating Other electricity

1200 1000 t CO2-eq

4760

800 600 400 200 0 1

5

9

13

17

21

25

29

33

37

41

45

49

Fig. 4. Weekly GHG emissions in Parikkala in 2009.

3.2. Weekly GHG emission estimates Figs. 3–5 present weekly variation of emissions in three different municipalities: Helsinki, Parikkala and Kuhmoinen. Helsinki (Fig. 3) is the capital city with 580 000 inhabitants. In Helsinki, the majority of emissions are from district heating by coal and natural gas, and the weekly emissions vary largely

depending on heating demand. On the contrary, the emissions of building-specific and electric heating are small, as more than 90% of buildings are connected to district heating. Emissions from ‘other electricity use’ are also higher than average in Helsinki due to presence of services and business. The emissions from other electricity consumption are higher in winter

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

than summer time for several reasons. Firstly, the emission intensity (g CO2/kWh) is more than double in winter than in summer time (127 g CO2/kWh in July 2009 and 263 g CO2/kWh in December 2009). Secondly, demand for lighting is notably larger in winter than summer months. In December, there is daylight for less than 6 h in Helsinki, whereas in mid-summer there is 19 h of daylight. Thirdly, ‘other electricity use’ includes also part of heating, as ‘electric heating’ covers only buildings the main heating system of which is electric heating. In particular in newer buildings, bathrooms are often equipped with electric floor heating; furthermore, electricity use of heat pumps is included in ‘other electricity’. In Helsinki, emissions from road transportation are relatively stable throughout the year. Emissions from agriculture are negligible, and emissions from waste management are also small. Parikkala is a municipality of 5900 inhabitants in South-Eastern Finland. There is less variability in emissions throughout the year than in Helsinki due to the importance of agriculture and road transportation. These emissions are at their highest in summer weeks, when emissions from heating and electricity consumption, in turn, are at their lowest (Fig. 4). Kuhmoinen is a municipality of 2600 inhabitants, in which there are more summer cottages than residential buildings. This is

800 700

Agriculture

Waste

Distric heating

Building-specific heating

Electric heating

Other electricity

Road transportation

600

400 300 200 100 0 1

5

9

13

17

21

25

29

33

37

41

45

49

Fig. 5. Weekly GHG emissions in Kuhmoinen.

clearly visible in the emission profile of road transportation (Fig. 5), as the emissions are high in particular during summer and other vacation periods. There is no district heating network in Kuhmoinen, and therefore the emissions from building-specific heating are higher than average. 3.3. Advance annual GHG emissions At the end of each year, the weekly GHG statistics of CO2report constitute the advance annual emissions of a municipality. The accuracy of the advance emission estimates, when compared with final statistics, depends on the sector and the circumstances in the municipality. For 42 out of 64 municipalities, the 2009 emissions from district heating were estimated before the 2009 district heating statistics of Finnish Energy Industries (2010a) were available. (The rest of the municipalities joined CO2-report after 2009 statistics were published.) In two of these 42 municipalities, there is no district heating network, whereas for 15 of them, information on the fuel use for district heating was obtained directly from the district heat producer, either at monthly or annual basis. In Fig. 6, the advance district heating emissions, estimated in CO2-report for 2009 based on 2008 fuel mix, are compared with the emissions from district heating calculated based on final 2009 district heating statistics for 22 municipalities (for 3 municipalities, the final 2009 statistics were not complete and therefore the comparison could not be made). The accuracy is best for municipalities with an isolated district heat network (no imports or exports of heat outside the municipality) using single fuel (for example natural gas or oil) for district heat production. Largest differences between the advance and actual emissions are found in municipalities, for which majority of heat is provided by an industrial producer. In municipality no. 9 in Fig. 6 the majority (82%) of district heat in 2008 was waste heat from an industrial plant, but the share dropped to 36% in 2009, and the remaining heat was produced with heavy fuel oil. By using the fuel mix of 2008, the 2009 emissions were underestimated. On the contrary, in municipality number 12, the industrial plant supplying district heat used twice as much wood in 2009 compared to 2008, and respectively less peat, heavy fuel oil and coal. Thus the advance emission estimate was higher than the emission estimate based on the final statistics. Also in the case of municipality number 19,

450000 Emission estimated based on final statistics Advance emission estimate

400000 350000 300000 t CO2-eq

t CO2-eq

500

250000 200000 150000 100000 50000 0 1

2

3

4

5

6

7

8

4761

9 10 11 12 13 14 15 16 17 18 19 20 21 22

Fig. 6. Comparison of emission estimates of district heating calculated in advance and those calculated based on final 2009 statistics in 22 municipalities.

4762

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

more peat was used in 2009 than in the previous year. In the municipality number 14, the district heat supplier reported distribution losses of 1% in 2008 but losses of 13% in 2009. Even though the advance estimate of heat demand was relatively accurate, the impact of losses increased the actual emissions. In the case of building-specific heating, there are no municipality-level statistics available, and therefore the advance emission estimate cannot be compared with final statistics. The CO2-report advance emission estimate is based on the latest information on municipal building stock, and final building statistics are updated to the CO2-report when available. However, the annual changes in building stock are usually small. The advance estimate of electricity demand is based on changes in heating demand for electric heating, and on the variability of other electricity consumption on weekly basis in the entire country (Finnish Energy Industries, 2010b). The advance estimate of electricity emission factor is also based on the monthly statistics of Finnish Energy Industries (2010b) and therefore, the accuracy of advance estimate of emissions from electricity consumption largely depends on the accuracy of the advance data of Finnish Energy Industries. In Fig. 7, the advance emissions of road transportation for 2009 are compared with the results of LIISA model (VTT Technical Research Centre of Finland, 2010). The advance emission estimates of CO2-report tend to overestimate emissions, as the development of car fleet towards less emitting vehicles and the impact of biofuel blends are not included in the advance estimates, which are based on changes in the traffic volume. The largest relative differences are found in points 43 and 51, which are municipalities through which a new highway was opened in 2009. It seems that the calculation model of CO2-report has overestimated the impact of the new highway on emissions. However, also the LIISA model to which the CO2-report results are compared is a calculation model with notable uncertainties. It appears that the LIISA model has not taken into use the latest measurements of transport volumes at these measuring points. In the case of waste management and agriculture, the advance emissions are the same as the emissions of the previous year, as there are no indicators available to estimate the change. In most cases, there is no significant year-to-year variability in these emissions. For example, in the agriculture sector, in half of the municipalities included in our study, the difference between annual emissions of 2008 and 2009 was smaller than 2%. However, if structural changes occur in agriculture sector in the

municipality, the differences can be relatively large. Such changes can be incorporated in the advance emission estimates if information is available. 3.4. Uncertainties in annual emission estimates of municipalities Uncertainty in the annual GHG estimates of municipalities is introduced in three different steps (see also Table 1):

 Uncertainty in municipality level statistics and own data collection.

 Uncertainty introduced in data processing, including assumptions made.

 Uncertainty in the emission factors and parameters of the Finnish GHG inventory and in their representativeness at the municipal level. The uncertainties in emissions from district heating are due to uncertainties in amount of fuels used and the related emission factors. The uncertainty in the amount of fuels used is of the same magnitude as in Finland in total, which vary from 71% for natural gas to 75% for peat and 720% for biomass, expressed as upper and lower bounds of 95% confidence interval (Monni et al., 2004; Monni and Syri, 2003). However, a possible bias in the reported fuel use by a single combustion plant (for example due to human error) would have a greater impact at municipality level than at national level. The CO2 emission factors used at municipal level are those used in Finnish GHG inventory, and therefore the uncertainties are also comparable (between 71% and 5%). However, in municipalities where waste is combusted to produce district heat, uncertainties in CO2 emissions are higher than in other municipalities, as the waste composition varies and is not accurately known. The CH4 and N2O emission factors of district heating boilers in Finland are estimated to have uncertainty ranges of 750% and  70% to þ150%, respectively (Monni et al., 2004). However, the importance of these emissions is small. The electricity consumption of municipalities is obtained from the statistics of Finnish Energy Industries (2010c), based on information from electricity distribution companies. Therefore, the uncertainty in electricity consumption by municipality is estimated to be small. The emission factor used is a national electricity emission factor calculated based on CO2 emissions from fuels used for electricity generation, and on electricity consumption in Finland (Finnish Energy Industries, 2010b;

600000 Emissions from LIISA model Advance emission estimate

500000

t CO2-eq

400000

300000

200000

100000

0 1

5

9

13

17

21

25

29

33

37

41

45

49

53

Fig. 7. Advance emission estimate of road transportation (CO2-report) for 2009 compared with the results of the LIISA model for 56 municipalities.

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

Statistics Finland, 2011). The division of electricity consumption into electric heating and other consumption introduces uncertainty in the results. Even though total electricity consumption in the municipality is well known, uncertainty arises due to the fact that different emission factors are used for the two types of electricity consumption (as electric heating is used more in winter months when emissions are higher). As the total uncertainty of CO2 emissions from energy industries in Finland is estimated at 73%, the uncertainty in the CO2 emissions from electricity consumption in municipalities can be estimated to be somewhat higher, around 75%. Uncertainty in CH4 and N2O emissions is higher, but their share of CO2-equivalent emissions of electricity consumption is only about 1%. Emission estimates from building-specific heating can be considered a major source of uncertainty in the municipal GHG estimates. The uncertainties are particularly important for municipalities, in which significant share of buildings are heated with building-specific heating and in which agriculture and waste sectors are not significant emission sources. Firstly, the statistics on floor space by building type and heating system are not always up-to-date, even though statistics are compiled annually. For example, if the heating system of an existing building is changed, the change is reflected in the statistics only if the change required a building permit. Secondly, the classification of heating system in which an oil boiler provides heat for more than one building may in some cases be classified as district heating and in others as building-specific heating, thus causing possible bias in the estimates. This is evident for some municipalities, which do not have a district heating network, but according to buildings statistics, have buildings which are classified as being heated with district heating. However, the main uncertainty arises due to the use of average energy demand (kWh/m2) for different types of buildings. This assumption does not take into account the age, construction material, design and use of buildings, thus introducing uncertainty into results. The wood combustion in buildings is based on statistics compiled by Finnish Forest Research Institute at an interval of about ten years. Therefore, it is evident that the information cannot capture inter-annual changes in the use of wood fuel. Furthermore, the statistics are based on a survey on households, which includes uncertainty as the amount of wood combusted is usually not accurately known. The emissions from road transportation are taken from the LIISA model (VTT Technical Research Centre of Finland, 2010) at municipality level. LIISA model is based on national fuel consumption, which is divided to municipalities based on, for example, the lengths of different types of roads within the municipality, and on traffic counts. Uncertainties of the LIISA model are not reported. However, it is likely that the uncertainties at municipal level are notably higher than at national level, because at the country-scale, sales of fuels are accurately known. In the agriculture sector, the emissions from enteric fermentation and manure management are estimated by using municipality level data on animal numbers and emission factors of the Finnish GHG inventory. Therefore, the uncertainties at municipality level can be considered similar as at national level, i.e. between 720% and 30% (Monni et al., 2006). In CO2-report, it is assumed that the division of manure management systems in the municipalities is the same as in Finland in average. This introduces some further uncertainty in the results at the municipal level. The uncertainty of N2O emissions from agricultural soils is dominated by the uncertainty in the emission factor. In Finland the uncertainty is estimated at  50% to þ 70% (Monni et al., 2006). At the municipality level the uncertainty is higher, in particular due to the assumption on the average use of N

4763

fertilizers per hectare, whereas at the national level, fertilizer sales (assumed to be equal with consumption each year) are well known. However, as the uncertainty is dominated by the uncertainty in the emission factor, the uncertainty at municipal level can be assumed to be of the same order of magnitude than at national level. In the waste sector, uncertainties in the CH4 emissions from landfills, calculated using the first order decay method, are high. In Finland in total, the emissions from landfills are estimated to include an uncertainty of about 730% (Monni et al., 2004). However, the uncertainty at the level of municipalities is estimated to be higher, due to lack of data on waste composition and amount of waste disposed in historical years. Even though similar data challenges are also faced in the compilation of national GHG inventory, their impact can be higher at the level of a single landfill, than when the emissions are estimated for all Finnish landfills at once. In addition, lack of information on the closed landfill sites within a municipality may be a source of bias. The emissions from wastewater treatment are based on information from VAHTI database and for uncollected wastewater, on the amount of population in rural areas. As the estimation method and data sources are the same as used in the national GHG inventory, the uncertainties at municipal level are estimated to be similar to those at national level, i.e. from  60% to þ 120% (Monni and Syri, 2003). However, possible errors in the data from VAHTI database would introduce a higher relative bias in the estimates at municipal, than at national, level. In total, the uncertainty in the Finnish GHG inventory (without land use, land use change and forestry sector) has been estimated to vary between  4% and þ7% in recent years (Monni et al., 2004; Statistics Finland, 2010a). At the level of the municipalities, the uncertainties strongly depend on the importance of different sectors. In cities, more than half of the emissions are from district heating and electricity consumption, which are well-known sources. For example in Helsinki, the share of these sources is 75%. In this case, it could be estimated that the uncertainty is of similar magnitude as in Finland in total. At the other end of ¨ the range are small municipalities such as Hameenkoski and Kiuruvesi, the emissions of which are dominated by agriculture and road transportation. In these municipalities, the uncertainties are notably higher.

4. Discussion In this paper we presented the CO2-report calculation model, which offers three types of information on municipality level GHG emissions: (1) weekly GHG emission estimates; (2) advance annual GHG estimates; and (3) final annual GHG estimates for comparison of municipalities and monitoring progress towards emission reduction targets. The annual GHG emissions of the municipalities included this study varied from 4.8 to 13.3 t CO2-eq/cap in 2009. Even though methodological differences hinder comparison of local GHG inventories carried out in different studies, it is interesting that the range of per capita emissions in Finnish municipalities is similarly wide as can be found in comparisons between very different city regions in developing and industrialized world, such as between Jakarta, Mexico City, Tokyo and Los Angeles (Sovacool and Brown, 2010). Our results are in accordance with the finding that per capita emissions are in developed countries lower in cities than the national average (e.g. Dhakal and Shrestha, 2010; Hoornweg et al., 2011): in our sample of Finnish municipalities, the emissions were higher in small municipalities than in larger cities. Yet the large per capita emissions in small Finnish municipalities are

4764

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

often due to through traffic or agricultural activity. These emissions do not reflect the emissions caused by the local citizens and are also difficult to tackle with local measures. Compared to other studies of large cities (Kennedy et al., 2009; Sovacool and Brown, 2010), the per capita GHG emissions in the largest cities of this study, Helsinki, Espoo and Tampere, are in the lower end at 4.9–5.2 t CO2-eq, with only Barcelona and London at the same level. Western cities with roughly similar climate conditions, such as Denver, Toronto and Prague, had two-threefold larger per capita GHG emissions. In a recent review of 100 cities (Hoornweg et al., 2011), the pattern is similar: out of Western cities in roughly similar climate conditions only Oslo and Stockholm had lower per capita emissions than Helsinki, Espoo and Tampere. Main reasons to this good performance of Finnish cities are relatively small emissions from surface transportation and the low emission level of average electricity consumption in Finland (196 g CO2/kWh in 2009), together with extensive district heating produced by CHP used in all these Finnish cities. Our study excluded industry, but its inclusion would not have had an impact on this conclusion, as industry is not a major emitter in any of the three cities. Comparisons of the advance emission estimates with that of final statistics show that the accuracy of the preliminary estimates is generally good, but it varies by municipality and sector. The accuracy is best for municipalities in which significant annual changes, for example in agriculture or district heat fuel mix, do not occur. In those cases, the indicators used to compile advance estimates are sufficient to reflect the inter-annual changes. In the case of municipalities with significant changes, the accuracy of the estimates is the better the more information on the changes is available. For example, the district heating suppliers in 15 municipalities provide fuel mix information for CO2-report in a timely manner, enabling good accuracy of advance emission estimates. When analyzing the emissions of municipalities and their reduction potentials, it is important to distinguish between those factors that can be directly addressed with local measures and those, which require national-level measures. Local decisionmakers can influence emissions from local transport by land use planning and by providing public transportation and improving the possibilities of biking and walking. A dispersed spatial structure typically increases substantially car traffic in Finland. The possibilities of municipalities to affect through traffic are very limited. The results show that small communities along the main highways can have manifold per capita emissions from transport compared to similar communities elsewhere. The reduction of these emissions requires national-level measures. The emissions from electricity consumption can be reduced by energy saving and energy efficiency measures. The local authority can realize energy saving measures in its own buildings and promote the energy saving measures of other local actors by e.g. information campaigns. Municipality-level investments in renewable electricity generation (often CHP in Finland) reduce the emissions from electricity production. However, with the calculation methodology used here this is not directly visible in the emissions of the municipality in concern, it only has a minor effect on the average emission coefficient of the national electricity production. There are methodologies available to account for local small-scale electricity production while using the national emission factor for electricity, such as the method of the Covenant of Mayors (European Union/Covenant of Mayors, 2010). Perhaps the sector where municipality-level measures have the largest impact in Finland is district heat consumption. In Finland, the municipalities which are active in climate change mitigation most often address district heat production with their

own measures (Savikko, 2009). They typically expand the district heating networks and increase the share of biomass fuels in their heating boilers and CHP facilities. For example the municipalities ¨ aki, ¨ ¨ ol ¨ a, ¨ Padasjoki and Eurajoki moved to mainly of Mynam Kark biomass-based district heating during the years 2008–2009. Also in many other small municipalities in Finland district heating is produced mainly by wood chips and expansion of production is planned. The municipalities also have a role in providing information and removing barriers from the switch from fuel oil to renewable energy in building-specific heating. Present methodologies for local and national GHG inventories cannot completely quantify the emission reduction occurring due to changes in agricultural practices—the estimated emissions are only affected by change in, for example, animal numbers or cultivated area. Several municipalities, for example in Kainuu in Northern Finland promote organic farming, and information on its GHG impacts would be of interest for local decision makers. However, accounting for emission impacts of different agricultural practices in the circumstances of Finland would require further scientific research. In Finland, municipalities are responsible for waste management. The emissions from waste management should be reduced, in the first place, by reducing waste generation, and by recycling and reuse. However, significant amounts of waste are disposed to landfills. Emissions from landfilling can be reduced by biological treatment of organic waste, and by landfill gas recovery. Waste incineration with energy recovery is an option to reduce emissions both from landfills and from district heat production, in particular where peat or oil are used in district heat boilers. As waste disposed to landfills emits methane for tens of years after disposal, the mitigation options in the waste sector do not immediately cause significant reductions in local emissions. Changes in waste treatment emissions can be shown by carbon footprint calculations, which determine the emissions of each ton of waste treated over its entire life cycle.

5. Conclusions The CO2-report provides weekly municipality-level information on GHG emissions, which constitute advance annual emissions at the end of the year. Final annual GHG estimates, for comparison of municipalities and monitoring progress towards emission reduction targets, are calculated when the official statistics become available, usually at the end of the following year. A key issue, which enabled the development of CO2-report is the high quality and availability of municipality level statistics in Finland. The availability of statistics on, for example, electricity consumption, agricultural activity, road transportation and district heat production allowed consistent calculation of emissions across municipalities. Uncertainties in the emission calculations are due to uncertainty in the underlying scientific methods and uncertainty in the statistical data used. Scientific uncertainty, for example in the emission factors for the agriculture sector, cannot be reduced by municipality-level action, whereas statistical data could be improved. The main area of improvement of municipality level data would be the energy use of oil heated buildings. In the currently used methodology, faster than average changes in the energy efficiency of buildings, or installation of building-specific renewable energy equipment such as solar heaters or heat pumps are not reflected in the emission calculations. Even though the year-to-year changes of the building stock are usually small, they are expected to accelerate in the future due to tighter building regulations and financial incentives provided at the national level

S. Monni, S. Syri / Energy Policy 39 (2011) 4755–4765

for the improvement of energy efficiency and building specific use of renewable energy. Therefore, gathering more information on oil use in buildings, for example by detailed surveys would be of great importance in order to reduce the uncertainty in the municipal GHG emission estimates. The calculation system currently covers 64 municipalities and 58% of Finnish population. This allows detailed comparisons, which were not feasible before, when the emissions of each municipality were calculated separately, and differences occurred due to different data sources, methodologies and assumptions. Per capita comparisons help to put the emission numbers into perspective, as municipal authorities and decision makers are not necessarily familiar with GHG emission inventories. Sector-bysector comparisons also allow the authorities to focus on sectors in which the emissions are higher than in other municipalities with similar conditions, as this often implicates that there is potential for emission reductions. The visual map presentation (at www.co2-raportti.fi) of municipalities involved also gives incentives for local authorities to participate, and publicly available comparison as such may motivate municipalities to take emission reduction action. The weekly GHG estimates are currently displayed at the web pages of the municipalities, and the service has awakened a lot of general interest. It has proven to be a useful tool for visualizing the GHG emissions and their causes for both citizens and local decision-makers. References Council of the European Union, 2007. Presidency Conclusions, Brussels, European Council 8/9 March 2007. 7224/1/07, REV 1, Brussels, Belgium, 2007. Dhakal, S., Shrestha, R., 2010. Bridging the research gaps for carbon emissions and their management in cities. Energy Policy 38, 4753–4755. European Union/Covenant of Mayors, 2010. How to Develop a Sustainable Energy Action Plan—Guidebook. Part II, Baseline Emissions Inventory. Publications Office of the European Union. Finnish Energy Industries, 2010a. District Heating in Finland 2009. Finnish Energy Industries. ISSN 0786-4809. Finnish Energy Industries, 2010b. Monthly Electricity Statistics (on-line data). /http://www.energia.fi/en/statistics/rapidreportS. Finnish Energy Industries, 2010c. Municipalities according to consumption. /http://www.energia.fi/fi/tilastot/sahkotilasto/kaytto/ kunnatsahkonkaytonsuuruudenmukaanS. Finnish Transport Agency, 2010. Automatic traffic measurements (on-line data).

4765

Graus, W., Worrell, E., 2011. Methods for calculating CO2 intensity of power generation and consumption: a global perspective. Energy Policy 39 (2), 613–627. Hoornweg, D., Sugar, L., Gomez, C., 2011. Cities and greenhouse gas emissions: moving forward, Environment and Urbanization. Published online 10 January 2011. doi:10.1177/0956247810392270. Kennedy, C., Steinberger, J., Gasson, B., Hillman, T., Havra´nek, M., Hansen, Y., Pataki, D., Phdungsilp, A., Ramaswami, A., Villalba Mendez, G., 2009. Greenhouse gas emissions from global cities. Environmental Science and Technology 43, 7297–7302. Kennedy, C., Steinberger, J., Gasson, B., Hansen, Y., Hillman, T., Havranek, M., Pataki, D., Phdungsilp, A., Ramaswami, A., Villalba Mendez, G., 2010. Methodology for inventorying greenhouse gas emissions from global cities. Energy Policy 38, 4828–4837. Kuittinen, V., Huttunen, J., Leinonen, S., 2010. Finnish biogas plant register no. 13 [Suomen Biokaasulaitosrekisteri n:o 13] (in Finnish). Publications of the University of Eastern Finland—Reports and Studies in Forestry and Natural Sciences. Ministry of Employment and the Economy, 2008. Long-term Climate and Energy Strategy. Government Report to Parliament. Ministry of Employment and the Economy Publications, Energy and Climate 36/2008, Helsinki, Finland (in Finnish). Monni, S., Syri, S., 2003. Uncertainties in the Finnish 2001 Greenhouse Gas Emission Inventory. VTT Tiedotteita—Research Notes 2209. Espoo 2003. Monni, S., Syri, S., Savolainen, I., 2004. Uncertainties in the Finnish greenhouse gas emission inventory. Environmental Science & Policy 7, 87–98. ¨ a, ¨ P., Regina, K., 2006. Uncertainty in agricultural CH4 and N2O Monni, S., Peral emissions from Finland—possibilities to increase accuracy in emission estimates. Mitigation and Adaptation Strategies for Global Change, 2006. Motiva, Oy, 2010. Normalization of heat consumption in buildings. [Rakennusten ¨ lammitysenergian kulutuksen normitus.] (in Finnish). ¨ a, ¨ J., 2007. Kasvener. A model for calculation of GHG emissions and energy Petaj balance at municipal level [Kuntatason kasvihuonekaasu-ja energiatasemalli]. Finnish Environment Institute (SYKE). Ramaswami, A., Hillman, T., Janson, B., Reiner, M., Thomas, G., 2008. A demandcentered, hybrid life-cycle methodology for city-scale greenhouse gas inventories. Environmental Science Technology 42 (17), 6455–6461. Savikko, R., 2009. About climate policy in Finnish municipalities. Questionnaire of the Association of Finnish Local and Regional Authorities in summer and fall 2009. [Ilmastopolitiikasta Suomen kunnissa. Kuntaliiton kysely ilmastopolitii¨ a¨ ja syksylla¨ 2009.] (in Finnish) Final report 15 kasta Suomen kunnissa kesall December 2009. Statistics Finland, 2009. Energy Statistics. Yearbook 2008. Statistics Finland, 2009. Statistics Finland, 2010a. Greenhouse gas emissions in Finland 1990–2008. National Inventory Report under the UNFCCC and the Kyoto Protocol. 25 May 2010. Statistics Finland, 2010b. Buildings. [Rakennuskanta] (in Finnish). Statfin statistical database. Statistics Finland, 2011. Energy Statistics. Yearbook 2010. Statistics Finland, 2011. Sovacool, B., Brown, M., 2010. Twelve metropolitan carbon footprints: a preliminary comparative global assessment. Energy Policy 38, 4856–4869. VTT Technical Research Centre of Finland, 2010. LIISA 2009. Road traffic exhaust emissions. /http://lipasto.vtt.fi/liisae/index.htmS.