Introduction

Buildings and construction contribute almost 40% to worldwide energy and process-related CO2 emissions (IEA, 2019), so there is no doubt that significant changes are needed in building sector in the view of the target of carbon neutrality by 2050 (European Comission, 2019). In much of Europe, the building stock is relatively old, for example, 50% of Finnish buildings were constructed before 1980 (Statistics Finland, 2017), and with low annual building stock renewal rates, around 1.5% in Finland (ROTI-panel, 2019), the renovation of existing building stock and improvement in the energy efficiency of old buildings offer an attractive solution to reducing emissions. In Finland, mostly due to its Northern location and weather conditions, heat consumption of buildings is considerable, totaling 26% of the final energy consumption of the country (Statistics Finland,  2019). The present study uses Finland as a case example, and as the weather conditions in Finland can be compared to the ones in other parts of Scandinavia, Baltic region, Western Russia, Northern America, and some parts of mountainous Asia, the results are relevant also outside Finland.

While there are acts and decrees considering the energy efficiency of a new building and the energy efficiency improvements in a renovation process, the owner of the building is eventually the one responsible for the investment in the energy efficiency improvements in the building. In Finland, municipalities are eminent building owners. The value of publicly used buildings in Finland is 45 billion euros and it represents 9% of the value of the whole building stock of the country (ROTI-panel, 2019). The extent of potential impact makes public buildings a promising research area. The municipalities also have more regulated processes than an individual building owner, which gives the decision-makers more power to have an impact on the choices made.

There is considerable unutilized potential in the public buildings regarding energy efficiency improvements and renovation. The majority of municipality-owned buildings in Finland are from the 1960s and 1970s, reflecting the rapid urbanization of Finnish society at the time (ROTI-panel, 2019). Like other European countries, Finland is also seeing a trend of movement to large urban centers coupled with aging population. Consequently, many municipalities are unable to maintain their buildings in good working order. At present, the maintenance backlog of public buildings in Finland is estimated to be 9 billion euro, and complete renovation would require 16.5 billion euro (ROTI-panel, 2019).

Energy efficiency of a building can be improved in multiple ways. Many of the technologies improving the energy efficiency of buildings, such as heat pumps and solar PV panels, have reached the Technology Readiness Level 9 and their cost has decreased rapidly in recent years. On the other hand, traditionally renovation of the building envelope has been favored in the northern conditions of Finland. Insulation and the integration into buildings of on-site energy production technologies are not commonly seen as two options but as two process steps following each other (Charles et al., 2019; Kuusk et al., 2014), or only renovation of the building envelope is considered (Bonakdar, 2018). However, there are multi-object optimization tools that are able to address this bifurcation with in-depth calculations. Niemelä (Niemelä, 2018) assessed cost-optimal renovation methods for four building types: brick and concrete large panel apartment buildings and educational and office buildings using multi-object optimization. The results showed that the most cost-effective main heating solution for the building types studied is a heat pump and that investments should focus on renewable energy production rather than renovation of the building envelope. Furthermore, Thygesen and Karlsson (Thygesen and Karlsson, 2017) concluded that it would be possible to build less insulated houses with the same final energy demand as more insulated houses if a ground source heat pump and heat recovery in ventilation are utilized, although the study included no assessment of the economic implications.

On the other hand, Kang et al. (2015) compared the effect of active and passive methods to the energy savings and concluded that passive energy efficiency improvement methods should be prioritized, while they contribute more to the energy savings. The passive methods studied by Kang et al. included methods that could not be used in renovating existing buildings, such as re-orientating the building. Also, the previous studies comparing active and passive methods are focusing solely on energy savings (Gracia & d., Navarrano, L., Coma, J., Serrano, S., Romaní, J., Pérez, G., Cabeza, L. F., 2018; Kang et al., 2015; Li et al., 2017), and the analysis is usually lacking an economical approach or it is reported inadequately. The energy savings point of view is usually based on an assumption that buildings use conventional energy; nowadays, it could be more economical to produce more energy by renewable energy sources than to focus on primary energy savings.

The first hypothesis underpinning the research is that energy efficiency improvements will bring the building owner savings in operating costs and can be profitable investments. Based on above-mentioned previous studies, a second hypothesis is also made that active energy efficiency improvements are more profitable than passive energy efficiency improvement measures. With the aim of assisting decision-makers to find optimal solutions for renovation of public buildings, the focus area of the work presented in this paper is the return on energy efficiency improvement investments. The specific objective of the study is to calculate the internal rates of returns (IRRs) of energy efficiency improvements in example public buildings in order to rank the alternatives for each building and determine possible savings in operating costs as well as the amount of investment required. The energy efficiency improvement investments considered are photovoltaic panels, ground source and air-to-water heat pumps, heat recovery in ventilation system, and renovation of some parts of the building envelope. The results of 12 buildings are analyzed to see if some building types or buildings of certain age have the most potential for savings. Also, the contribution of each active and passive improvement technology is assessed for case buildings. Thus, the novelty of this study is the profitability comparison between active and passive energy efficiency improvement investments in the context of renovation in multiple case buildings, and analysis of underlying reasons for the results. Assessment of savings in operating costs is also a complementary contribution to the research area.

The methodological approach used in the study is techno-economic modeling based on measured energy consumption data. Real energy consumption is used as a basis for energy demand simulation in the different scenarios. Though there are existing tools for optimizing the energy efficiency improvement investments, a new, lighter model is produced to form an understanding of the factors having greatest influence on the energy efficiency improvement investments in each case building. The aim is to generate the results with a compatible amount of information about a case building and to do it as reliably as possible. The most economically profitable improvements and the buildings with the highest savings potential can thus be chosen, and after that, more in-depth planning of investment can be done regarding each building. The model can be used to choose the investments in an earlier stage of investment planning, when compared to the existing models.

The study investigates 12 case buildings in the Finnish cities of Lappeenranta and Imatra in the South Karelian region. The case buildings can be divided into four types: office buildings, commercial buildings, schools, and industrial buildings. More information about the buildings is provided in the “Background data of the case buildings and calculation inputs” section. The studied buildings include both recently built and older ones, and they have different properties and design. The results of this study can be considered providing representative averages for different building types in Northern climate conditions.

The rest of the paper is organized as follows. “Energy efficiency improvements” section describes the technologies considered in this study. Background information about the case buildings is presented in “Background data of the case buildings and calculation inputs” section. “Techno-economic calculations” section presents the calculations and “Results” section concludes the results. “Discussion” section presents discussion and the final part of the paper, “Conclusion” section, conclusions.

Energy efficiency improvements

In this work, the term active energy efficiency improvement is defined as incorporating energy-efficient energy technologies into the building, for example adding a heat pump or solar photovoltaic panels to a building, or changing an existing heat recovery unit (HRU) in the ventilation system to a more efficient one. Active energy efficiency improvements are using or producing energy, and they are an active part of the systems in the building. Reduction of the thermal transmittance of the structure, e.g., by adding insulation or renewing some parts of the building envelope, is considered passive energy efficiency improvement.

It should be noted that with exception of a few building types, Finnish legislation requires that the energy efficiency of a building must be improved in Finland when the building is renovated, if it is technically, functionally, and economically feasible (1999, 132.Maankäyttö ja rakennuslaki, 117g § Energiatehokkuus, 1999). The regulations set limits for building block specific values such as the value of thermal transmittance of the envelope materials, for technical system parameters such as the efficiency of the heat recovery, and for overall energy consumption (kWh/m2). The energy efficiency improvements can be executed in multiple different ways, if the given targets are met. (Ministry of Environment, 2013).

Heat pumps

Heat is distributed in all of the case buildings by radiator networks using water as a working fluid. Thus, the present study focuses on ground source heat pumps (GSHP) and air-to-water heat pumps (AWHP). Exhaust air heat pumps could also be an option, but are omitted from the study because such pumps cannot supply the major part of a building’s heat demand as the heat source is limited. Furthermore, efficient heat recovery in the ventilation system means that there is little heat left for utilization by the exhaust air heat pump.

As Finland has a Northern climate, the outdoor air temperature level varies greatly, from − 30 to + 30 °C, while the temperature in the ground stays stable throughout the year, at about + 5 °C. In Finnish conditions, air source heat pumps have significantly lower COP values in wintertime than ground source heat pumps because of the cold outdoor temperatures. Air source heat pump units stop working at specific temperature conditions, and in winter, the unit needs additional electricity in order to melt ice building up due to condensation. The other parameter affecting the COP of the heat pump is the temperature of the radiator network. The output power of a heat pump is higher when the temperature level of the radiator network is lower, because the temperature difference is smaller.

Figure 1 presents the theoretical COP as a function of the temperature difference between the heat source and the radiator working fluid. Real COP values in heat pumps are always lower than the theoretical COP. The function for COP of the GSHPs used in this study was obtained from data for 14 different high temperature heat pumps on the market for the temperature lift range of 20 to 140 °C (Arpagaus et al., 2018). The values in the temperature lift range of 20 to 80 °C are on average 45% of the theoretical COP. Data for COP values of AWHPs were obtained from experimental measurements of one AWHP on the market (VTT, 2009) and real COP was calculated as approximately 25% of theoretical COP, Fig. 1.

Fig. 1
figure 1

Theoretical COP function with functions for GSHPs and AWHPs

Full size image

In the calculations, the costs for heat pumps and bore holes are average costs for new constructions in Finland. An AWHP system costs 380 €/kWth, of which 300 €/kWth is the cost for a heat pump (Haahtela & Kiiras, 2015) and 80 €/kWth for a condenser. Cost for the condenser is 60–100 €/kWth based on three offers, and thus an average price of 80 €/kWth is chosen for calculation. A GSHP system costs 1,064 €/kWth, of which 764 €/kWth is the cost for the bore holes (200 m, 40 W/m (Kukkonen, 2000)). The connection pipes to the boreholes cost 12.8 €/grm2. (Haahtela and Kiiras, 2015) The yearly fixed operating cost of the heat pumps is estimated to be 1% of investment cost for a GSHP and 2% for an AWHP (Paiho et al., 2017). The compressor of a heat pump must be renewed after 15 years of operation (Paiho et al., 2017), so in the economic calculations, an additional investment in a new heat pump appears in year 15.

The heat pump is dimensioned to cover the whole heat demand of the building for simplification of the calculation, although this is not the economically optimal size. The economically optimal size of a heat pump depends on climate, thermal load, and electricity price (Fischer et al., 2016), and in Finnish conditions, an additional heat source is usually used in addition to the heat pump.

Heat recovery

Heat recovery in ventilation systems is important because 27–55% of the heat losses of a building occur through the exhaust air flow (Seppänen & Seppänen, 1997; Virta and Pylsy, 2011). Heat recovery can be implemented, if air supply and exhaust air flow is assisted with fans, which is usually the case in larger buildings. If only exhaust air flow is assisted with a fan, an exhaust air heat pump would be an applicable option.

The efficiency of a heat recovery unit in a ventilation system is described by temperature efficiency, which means the capacity of the unit to recover heat measured in standard conditions (Seppänen, 1996). The temperature efficiency varies throughout the year as the outdoor air temperature varies, so the final efficiency of the heat recovery unit is given as yearly efficiency. Temperature efficiency for exhaust air is calculated as follows (Ministry of Environment, 2003):

$${\eta }_{\mathrm{T}}=\frac{{T}_{\mathrm{Extract}}-{T}_{\mathrm{Exhaust}}}{{T}_{\mathrm{Extract}}-{T}_{\mathrm{Outside}}}$$
(1)

All the case buildings have fan assisted ventilation and the systems are equipped with either regenerative heat recovery with a rotating wheel or indirect recuperative heat recovery with glycol as a working fluid. The general temperature efficiency of regenerative heat recovery systems is 60–80% and that of indirect recuperative heat recovery 45–60% (Seppänen, 1996). In renovated buildings, the yearly efficiency of heat recovery must be at least 45%, as mandated by the Ministry of Environment (Ministry of Environment, 2013).

The costs for different HRUs vary depending on the supply air flow, which determines the size of the ventilation unit. A recuperative HRU with glycol as a working fluid costs 12,000–33,200 €/piece for flow rates 2–12 m3/s, and the glycol pipes cost 147 €/m. A regenerative heat recovery wheel costs 19,100–30,500 €/piece for flow rates 4–13 m3/s. (Haahtela and Kiiras, 2015) Operational costs of ventilation are associated with cleaning of the ventilation ducts and changing the filters, which are part of the whole ventilation system. Those operations belong to normal operational costs of a building, and thus the yearly fixed operational cost for the HRU alone is assumed to be 0% of capex.

Solar photovoltaic panels

Solar PV panels were chosen for analysis rather than solar thermal panels, because of the maturity of the technology and because of the mismatch between the time of the greatest heating demand and the time of solar thermal peak production. The yearly sum of global irradiation meeting, an optimally inclined solar PV panel in South Karelia region, is 1,050 kWh/m2 (European Comission, 2017). A peak kilowatt can be obtained from a 5.5 m2 panel area (Canadian Solar, 2020).

Costs for PV panel systems have decreased by about 80% in the last decade (International Renewable Energy Agency (n.d) ), and prices in Finland are in the range of 0.8–1.05 €/W for a 10–100 kW roof-top PV system and 0.7–0.8 €/W for a 100–250 kW roof-top PV system (Ahola, 2019). The lowest prices in the range are used in the calculations due to on-going decrease of the prices. The annual fixed operating cost of a PV system is estimated to be 1.5% of capex including the replacement of the inverters (Kosonen et al., 2014).

Structural improvements

Thermal transmittance, U (W/m2K), is a value that describes how much heat a structure conducts. Specific heat loss is obtained from the thermal transmittance and the surface area. The heat loss varies with respect to the temperature difference between the two sides of the structure, inside and outside. Building regulations (Ministry of Environment, 2013) specify requirements for thermal transmittance after renovation, but as mentioned earlier, it is possible to decrease the energy demand of the building by other means and still meet the energy efficiency targets.

Whether the addition of insulation to the walls or renovation of other structure is more economically feasible, depends on the initial value of thermal transmittance, the share of heat loss conducted through a specific structure, and the cost of the materials, which are studied further regarding each building type in the “Background data of the case buildings and calculation inputs” section.

Background data of the case buildings and calculation inputs

Currently, the case buildings use district heating (DH) or a natural gas (NG) boiler as a heating source, mechanical ventilation and related heat recovery, and in most cases, a building automation system.

Both municipalities purchase renewable electricity for public buildings: in Lappeenranta, the municipality buys wind power, and in Imatra, hydro power is bought for all municipality-owned properties. The raw energy cost of the electricity price is estimated to be 38.65 €/MWh, which is the average of the electricity price in Finland for 2003–2018 (Nord Pool, 2018). The total cost of electricity is estimated based on actual invoices from buildings in Lappeenranta. The final electricity costs in the electricity tax class 1 without value added tax (VAT) comprises the raw energy cost, 40%, network costs, 37%, and energy tax, 23%. The total price of electricity is thus 96.63 €/MWh (VAT 0%).

The cost for district heating consists of energy, transfer and flow rate costs, and the flow rate share of the DH price depends on the heat demand of the building. The total price for district heating in the case buildings is in Lappeenranta 76.50–83.91 €/MWh (VAT 0%) (Lappeenrannan Energia Oy, 2020a) and 71.35–74.01 €/MWh (VAT 0%) in Imatra (Imatran Lämpö Oy, 2016). The price for natural gas is estimated using the lower heating value of NG as 71.37 €/MWh (VAT 0%) including excise tax (Lappeenrannan energia Oy, 2020b). The energy prices are presented in Table 1.

Table 1 The energy prices used in the study
Full size table

Background data for the case buildings is shown in Table 2. The division of the building types is as in the usage group division in the Decree of the Ministry of the Environment (Ministry of the Environment, 2017), apart from the building type “Industrial buildings,” which are listed as “Other buildings” in the Decree. The Decree gives information about heat loads and capacity usage of different building types. The values for temperature efficiency of heat recovery unit and heated net area of the building as well as later introduced thermal transmittances and areas of each existing structure used in the calculations were taken from the Energy Performance Certificate (EPC) of each building. A few of the buildings and their envelopes are protected because of their historical value, but according to Gabeza et al. (Gabeza et al., 2018), implementing active and passive energy efficiency improvements in historical buildings can be done successfully.

Table 2 Background information about the case buildings. Locations are Lappeenranta (LPR) and Imatra (IM)
Full size table

In addition to the information in Table 2, the mall, no. 7, is a mix usage building and has premises for residential purposes as well as for commercial purposes, which is taken into account in the calculations with typical heat loads and capacity usage from the Decree of the Ministry of the Environment. It should also be mentioned that the school no. 8 already has a PV system installed on its roof.

Hourly or monthly heat consumption data were obtained for the reference year 2017 or 2018 and electricity consumption data for 2017, 2018, or 2019 for each case building. The electricity demand is based on automatic meter reading measurements and the heat demand based on measurements made by the energy company. Table 3 shows if the data for each case building is obtained monthly (mo) or hourly (h), and the heat demand and electricity demand divided by the size of the building.

Table 3 Form of electricity and heat consumption data as well as the heat and electricity demand data
Full size table

Heat losses in a building result from ventilation, conduction through the building envelope and air leakage. Figure 2 shows the average shares of heat losses from ventilation and the building envelope for each building type. In the case buildings, heat loss through air leakage is minimal (0.005–0.035%) and is thus not presented in the figure.

Fig. 2
figure 2

Share of heat losses from the building envelope by structural surfaces and ventilation for each building type, averages for a 1-year time period. The values are calculated from the Energy Performance Certificate (EPC) of the case buildings

Full size image

The share of heat losses from ventilation and thermal transmittance through building envelope is roughly half and half, with the exception of school buildings, where heat loss from ventilation is considerably smaller due to their more recent construction and differences in energy efficiency legislation for such buildings. Thus, the potential in energy savings in ventilation is usually greater than renovation of one structural surface. When considering structural changes, the most beneficial structure to renovate can be selected using information about thermal transmittance and surface areas.

As can be seen in Fig. 2, in office buildings, thermal conduction through windows is responsible for the largest share of heat loss from building envelope, on average 23%. Heat losses through walls account for an average 9% of total losses through the building envelope, which is considerably less than losses through windows. In commercial and school buildings, the same trend is visible; the share of heat loss through windows is 18% in commercial buildings and 24% in schools. Heat loss through walls accounts for the second largest share of total heat losses from structural surfaces in these buildings. In industrial buildings, the largest share of heat losses is through the base floor, 15% on average, followed by losses through the roof, 12% and walls, 12%. The difference to other building types is due to the typically smaller area and number of windows in such buildings, and already good U-values. The low share of heat losses through the walls, roof, and base floor can be explained by the Finnish building legislation, which has required reasonable insulation in the building envelope from 1985 onwards (Ministry of Environment 1985).

In view of its important role in heat loss, window renovation was chosen as a passive energy efficiency improvement for office, commercial, and school buildings. The costs of renewing windows are building specific, as the windows and the surrounding structures differ for each building. For simplification, a general cost for window renovation was applied. Cost of demolition of old windows is estimated to be 7.2 €/m2 and the price for new windows 179.17 €/m2 (U = 0.9 − 1.0 W/m2K) (Rakennustieto Oy, 2014). Annual fixed operating cost was assumed to be 0% of capex, because renewal of the sealants is assumed to belong to normal operating costs of the building and the need for sealant renewal is not dependent on the window renovation.

In industrial buildings, the base floor causes the most heat losses. Adding insulation to the base floor would mean removing the existing floor, adding insulation and building a new floor. The second largest share of heat losses happens through the roof, followed by walls. For two of the three industrial buildings, adding insulation to the floor, the roof, or the walls would not be possible without considerable extra construction work. In one of the industrial buildings, adding insulation to the roof or to the walls would be possible if done at the same time as a major façade renovation. As these improvements would require other major renovations as well installation of the extra insulation, their cost is difficult to calculate. Thus, replacement of the windows was used as the passive energy efficiency improvement for all the building types.

Techno-economic calculations

Calculations related to energy investments and operational costs are presented in this section. Heat demand calculation per month is performed because consumption data is mainly in the form of monthly values. Electricity consumption is calculated hourly, so the share of on-site PV production can be estimated.

Theoretical heat demand and the effect of insulation and ventilation

First, the theoretical heat demand of the buildings is needed to be able to see the effects of utilization of a heat pump, replacement of the HRU, or renovation of the selected building structure. The theoretical monthly heat demand was calculated using applicable energy efficiency legislation and values in the EPC (Ministry of the Environment, 2017) as well as 10 years average outdoor air temperatures. The reference heat demands for the buildings are for 2017 or 2018. The outdoor air temperatures in Lappeenranta in 2017 and 2018 and the 10 years average for 2008–2018 are presented in Fig. 3.

Fig. 3
figure 3

Temperatures in Lappeenranta in 2017, 2018, and 10-year averages for 2008–2018 Finnish Meteorological Institute. n.d.

Full size image

2018 experienced more extreme temperatures than average for the period 2008–2018, whereas the temperatures of 2017 followed the 10 years average quite well, except for the warmer summer. As the average temperatures are used for calculating the theoretical heat demand, the result can differ from the reference values. The same weather data is used for buildings in Imatra, because the Lappeenranta measuring point is the closest one to Imatra.

The heat demand takes into account heat loads inside and outside the building (people, equipment, and the sun) and heat losses from ventilation, structures in the building envelope, and air leakage. Building-type specific loads, capacity usage, and time of use were taken from the applicable Ministry of Environment decree (Ministry of the Environment, 2017). The effect of sun through the windows and other parameters related to sun load calculation were calculated as per the energy certificate calculation instructions issued by the Ministry of Environment (Ministry of Environment, 2006). Heat demand for domestic hot water, DHW, was calculated using data for hot water demand in each building type (Ministry of the Environment, 2017) or the EPC, and the assumption was made that hot water demand varies throughout the year as in a residential building (Seppänen, 2001). The temperature of DHW was assumed to be 55 °C.

The theoretical heat demand was defined for three scenarios; the situation as is, following structural improvements, and following the replacement of the existing HRUs with units with maximum temperature efficiency for the corresponding technology (for regenerative HRUs 80% and for recuperative HRUs 60%). The results are presented in Fig. 4. For comparison, the reference heat demand based on measured values is also presented. The permissible difference between the calculated heat demand in situation as is and the real reference values was ± 10%, which was obtained by adjusting the indoor temperature, bearing in mind that the theoretical heat demand is calculated with average temperatures from 2008 to 2018.

Fig. 4
figure 4

Real heat demand and calculated demand for three cases, as is, with structural improvements and with more effective HRUs, presented as area specific heat demands

Full size image

As seen in Fig. 4, the real and theoretical demand of the mall (no. 7) differ more than 10%. The mall has apartments in addition to commercial spaces, which was taken into consideration in the calculations, but the heat demand in residential housing is more unpredictable and the difference to the theoretical values is likely due to residential usage.

The heat demand in the scenario with new windows compared to the heat demand in scenario as is depends on the initial values of thermal transmittance and number and size of the windows and the heat demand in the scenario with a new heat recovery unit depends on the heat recovery system type and its maximum efficiency. In most cases, heat demand is the lowest with a new heat recovery unit, although where there is already efficient heat recovery, the lowest heat demand occurs with new windows.

The calculated heat demand is distributed differently throughout the year than the reference values, which is likely due to normalization. In the monthly calculations, there is no space heating demand in the summer due to the monthly average temperatures, although in the real reference values, there is heat demand also during the summer. The cooling demand, i.e., negative values for heating demand are not taken into consideration in the model, while the electricity needed for providing the cooling via ventilation is part of the reference electricity demand.

Heat pump and the radiator network

Calculation of electricity demand of the heat pump was done for each month using the above presented heat demands and average temperatures of the heat source, which is either outside air or ground temperature. Monthly calculation of the heat pump operation was done to get an overview of the profitability of the heat pump investment and to address differences in heat pump operation in different seasons. Calculation of COP for the heat pump is based on the functions from Fig. 1 as follows:

$$\left\{\begin{array}{c}{COP}_{GSHP}=0.45\bullet {COP}_{\mathrm{Theoretical}}={99.259\bullet \Delta T}^{-0.902}\\ {COP}_{AWHP}=0.25\bullet {COP}_{\mathrm{Theoretical}}={55.144\bullet \Delta T}^{-0.902}\end{array}\right.$$
(2)

The calculation was done separately for space heating via the radiator network and for domestic hot water heating, because of their different temperature levels. It was assumed that all the space heating is distributed via the radiator network, and the supply air heating was not calculated separately. Because of the lower temperature levels in supply air, this assumption is conservative and the real electricity consumption is lower than calculated.

The radiator networks in the case buildings are based on high temperature heat distribution, and the heat exchange area of the radiators is relatively small. Thus, the radiator network might need some renovation to meet the demands of a low temperature heat distribution network. The cost of such radiator network renovation is assumed to be included in the major renovation costs, so it is not treated further in this paper. The heat exchange area was optimized manually in the calculation so that the temperature in the heat distribution network in January does not exceed 50 °C. In the calculations, the working fluid temperature in the heat distribution network was calculated with the monthly average temperature.

The working fluid temperature was calculated based on the heat transfer from the radiators required to keep the building at the desired temperature. A radiator transfers heat to room air by radiation as well as by convection. The average temperature is calculated as a weighted average of the temperature of the convection heat transfer and the temperature of the radiation heat transfer with the shares of both heat transfer types. It was assumed that the buildings use one-plane radiators and heat transfer is 50% radiation and 50% convection.

$$\left\{\begin{array}{c}\frac{0.5\bullet Q}{d \bullet 24 \mathrm{h}}=A\bullet \varepsilon \bullet \sigma \bullet \left({T}_{\mathrm{s}}^{4}-{T}_{\infty }^{4}\right)\\ \frac{0.5\bullet Q}{d \bullet 24 \mathrm{h}}=A\bullet \alpha \bullet ({T}_{\mathrm{s}}-{T}_{\infty })\end{array}\right.$$
(3)

The calculated temperature, Ts, was assumed to be equal to the average working fluid temperature. The working fluid comes into a radiator at a higher temperature and leaves at a lower temperature after the heat transfer. Thus, the temperature that the heat pump needs to generate is higher than the average temperature of the working fluid, and it can be calculated with the working fluid’s temperature when it exits the radiator, which is now assumed to be equal to room temperature. In reality, there is always a temperature difference between the room air temperature and the working fluid temperature in heat transfer. Emissivity of the radiator was assumed to be 1 (Seppänen, 2001), and the convective heat transfer 4 W/m2K (VDI Verlag, 1988). The COP and the electricity used by heat pump can be solved with these values of monthly heat demands, working fluid temperatures and temperatures of the heat source.

The size of the heat pump is calculated using the highest heat demand based on monthly values. In the real situation, the heat pump would cover the major part of the load but not the whole load as in the calculations, so the profitability of a heat pump could be higher in more precise calculations.

Solar PV panel system

It was assumed that the whole roof-top area excluding any inlets or ducts can be fitted with solar PV panels, with 70% fill-up rate. The size of the PV plant does not need to be optimized precisely because the internal rate of return, IRR, does not significantly decrease if the plant is oversized (Simola et al., 2018). Thus, the size of the PV plant was roughly optimized manually to maximize the IRR, given the limitation of the available roof-top area. The hourly solar irradiation data used is for Lappeenranta and for panels with orientation to the south and a tilt angle of 15°. The data was generated with the energy modeling software Homer Beta 2.68. It was also assumed that the panels do not generate electricity in the winter months, from December to February, because of snow covering the panels. Degradation of the panels was accounted for using 0.5% decrease in yearly efficiency (Simola et al., 2018).

Because the heat pump calculations were made on a monthly basis, the electricity consumption of the heat pump was divided for every hour of the month. On-site consumption of the electricity produced by the PV panels and consequently the economic performance of the PV system might thus vary with more precise hourly calculations.

Economical assessment of improvements

Discounted net cash flow was used for economical assessment. IRR as well as payback time of the investment can be evaluated from the discounted net cash flow calculation. The IRR shows the economic profitability of an investment and allows the profitability of different investments to be compared. IRR is calculated by solving the following equation:

$$NPV=\sum\nolimits_{n=1}^{N}\frac{{I}_{n}}{{(1+IRR)}^{n}}=0$$
(4)

Investment and operational costs for each case building and the technology were calculated with cost information given in the “Energy efficiency improvements” and “Background data of the case buildings and calculation inputs” sections, and additional costs for planning and installation, together 20% of total investment, were used in the calculations. A 20-year time span was chosen for consideration. The interest rate can be moderate, because the investments are not very risky, so an interest rate of 5% p.a. was used for the discounted net cash flow. The IRRs were calculated using the baseline of the existing situation (as is). For the heat pumps, a scenario with PV panels in the system was also calculated to address their effect on the electricity price.

Results

The IRR for each investment is presented in Table 4 compared to the as-is situation.

Table 4 Internal rates of returns for active and passive improvements for each case building. For those cases where the efficiency of the HRU was already maximum for the technology or the U-value of the windows was already 1 W/m.2 K, the text “not feasible” is shown. “No solution” means the IRR cannot be solved because of negative cash flow
Full size table

When the size of the plant is roughly optimized, a PV system has an IRR of about 5% in almost all the cases studied, and IRR of 7% in those cases where the PV system size is over 100 kW and thus the price of the system is lower. Investment in GSHPs has an average IRR of 9% and AWHPs 14%. The IRRs are up to 1% higher for GSHPs when the effect of PV panels is taken into account. With AWHPs, the effect is even higher, on average 6%, due to the higher electricity demand of air-sourced heat pump. The increase in energy efficiency resulting from installation of a new HRU gives an average IRR of 57%, although the IRR varies greatly for different cases. Renewal of the windows has an average IRR of 2%. The feasibility differences can also be seen in the area specific investments: in the studied buildings, on average, the feasible investment for PV panels is 11 €/m2, for HRU improvements 17 €/m2, 54 €/m2 for GSHP and 15 €/m2 for AWHP investments, and 27 €/m2 for window renovation.

A parameter having considerable impact on the economics of investment in heat pumps is the electricity price. The results of sensitivity analysis of electricity price changes for one case building, library (no. 5), are shown in Fig. 5.

Fig. 5
figure 5

Electricity price sensitivity analysis for the library (no. 5). The price of electricity is changed ± 20% to show the effect of price changes on IRR of GSHP and AWHP investment

Full size image

As can be seen in Fig. 5, the price of electricity has a greater effect on investment in AWHPs. A 5% increase in the electricity price means that investment in AWHPs ceases to be profitable for this case building, but even with a 20% increase in the electricity price, the IRR of the GSHP would not drop significantly. The difference is due to the fact that GSHPs have an average seasonal COP, SCOP, of 3.9 whereas AWHPs have a SCOP of 1.8 in the case buildings, which means that GSHPs use less electricity to produce the same amount of heat energy as AWHPs. Thus, it can be claimed that a GSHP is a more risk-free investment as regards variation in the electricity price. Furthermore, for the library, the levelized cost of heat produced by an AWHP is about 67 €/MWh and by a GSHP about 58 €/MWh. For PV, the levelized cost of electricity is 83 €/MWh. When these are compared to the original costs for DH, NG, and electricity, it is clear why the renewable technologies are competitive options for the existing ones.

The IRR for GSHPs varies between 5 and 12% and IRR of AWHPs between 2 and 28%, excluding one result which could not be calculated (no value). The more significant variation in the economic performance of AWHPs is due to the price differences of the prior heating source and the greater effect it has on AWHPs than GSHPs, along with variation in room temperatures inside the buildings. AWHPs have lower initial investment costs than GSHPs, which makes them more profitable than GSHPs in most of the case buildings. The IRR for installation of a new HRU is higher when there is currently low efficiency and there are fewer ventilation units in the building. The IRR of window renovation is above 5% in only 2 of 10 case buildings. When comparing active and passive improvements, it can be seen that for these buildings, passive energy efficiency improvements are usually not as financially attractive as active improvements.

When the limit to profitability (IRR) is set to 5%, the most economically profitable energy efficiency improvements are chosen for each building based on IRRs in Table 4. The technologies are chosen only if they are technically feasible for the building. Figure 6 shows the potential yearly savings calculated with discounted cash flow per chosen technology as a percentage of the operating costs before the investments. Total realized savings if all the feasible technologies are implemented are presented as a separate column.

Fig. 6
figure 6

Net savings induced by chosen technologies for each case building, as well as total realized savings, if all the chosen technologies are implemented. The savings are calculated for the first year of operation. Savings slowly decrease over time because of the effect of PV degradation and interest rate costs

Full size image

In the analysis, PV panels are chosen for buildings which have high electricity usage and large roof area, building nos. 2, 6, and 12. Investment in a new HRU is chosen whenever feasible. A GSHP is chosen for those cases that have the necessary space on the property and do not have restrictions concerning with ground water protection. A GSHP is favored although an AWHP would be more economically feasible in regard of Fig. 5. If an AWHP is significantly more economically feasible than a GSHP, it is chosen for the case building even when a GSHP would be technically feasible, which is the situation for the logistical building no. 10 and industrial building no 12. For the rest of the buildings, an AWHP is chosen, if the IRR is over 5%.

The savings resulting from the studied investments were calculated for each technology separately and compared to the existing situation to illustrate the savings potential in terms of reduction in operating costs. In this case, operating costs are the price of heat and electricity used in the building. If multiple investments are made, the total savings must take into consideration the combined effect on operating costs of all the technologies, because savings resulting from one technology have an effect on the other technologies. For example, if the efficiency of the HRU is increased, the profitability of investment in a heat pump increases, because the heat pump does not need to generate as much heat and its size can thus be reduced. On the other hand, if the investment in a HRU is compared to the situation with a heat pump as a heating method instead of district heating, the savings resulting from the HRU investment decrease. Savings from a heat pump and investment in an HRU are duplicated, so the total realized savings would usually be lower in buildings where investments are made in both a heat pump and an HRU. A similar interaction can be found for investment in both window renovation and heat pumps. PV panels affect the price of the electricity used in a building, which has an effect on those investments of which profitability depends on their electricity usage, such as heat pumps. The reduction in the price of electricity means that, for example, the rate of return on investment in heat pumps increases. As a result of the relationships between the different investments, the total realized savings can be higher than the sum of savings for each technology, if investments were made in a PV system, or the total realized savings can be lower than the sum of savings for each technology, if investments were made in both a new HRU or window renovation and a heat pump.

As seen in the results, the energy efficiency improvements are building specific, and they depend on many variables. The sum of potential savings from adoption of the different technologies varies between 20 and 73%, and total realized savings between 20 and 59%. The total realized saving potential is on average 35% of the yearly energy-related operating costs. The largest savings result from usage of ground source heat pumps and improvements in heat recovery efficiency, Fig. 6. The difference between AWHPs and GSHPs is clearly visible in the analysis. In the studied buildings, installation of PV panels would bring the least benefit in terms of savings per year.

An analysis is made to see if the savings potential differs with the type of building, Table 5. Furthermore, the relation between the savings potential and the building year is presented in Table 5.

Table 5 The savings potential averages studied by building type and by building year
Full size table

When the buildings are divided into groups by building type, it can be seen that the most savings potential is in school buildings and commercial buildings, though the investments are higher than in office buildings and in industrial buildings. When the buildings are divided into groups related to their building year, it can be seen that there is only little differences between the savings potential in the groups. The saving potential in buildings of different ages approximates the overall average savings potential, 35%. The school no. 9 and industrial building no. 12 are calculated in the buildings built before 1980 although they have parts that are built in 2000s, which can level the differences between buildings of different age. The fact that the energy efficiency improvement techniques are individually chosen for each case building affects the results so that the buildings’ types or ages are not fully representative as generalized results. Nevertheless, it can be concluded that the savings potential is high despite the building type or the age of the building and the payback time of the investment is bearable.

Rose et al. (Rose et al., 2016) studied minor and major renovations in Estonian and Danish context and concluded that the investments needed for energy-related renovation were 120–170 €/m2. As the prices have decreased, this study concludes that the investment need for energy efficiency improvements is on average 60 €/m2 when all the buildings are taken into consideration. The investments in energy efficiency are minor when compared to the major renovation costs. Likewise, it is concluded by Tuominen et al. (Tuominen et al., 2013) that there is a significant energy savings potential in Finnish building sector with a minor increase in investments in construction and renovation.

Total saving potential for all 12 case buildings amounts to approximately 740,000 €/a for an investment of 4,572,000 €. The simple payback time for the investments would thus be 6.2 years. Some additional investments might be needed for actions such as replacing the heat recovery unit, because it would be easiest to renew the whole ventilation unit at the same time as replacing the HRU.

The analysis indicates that it is possible to reach the given targets for energy consumption in the Finnish legislation with active energy efficiency improvements alone, because the HRU decreases the energy consumption and the heat pump decreases the amount of purchased energy. The energy consumption reference is defined in the legislation for all building types other than industrial buildings, and case buildings’ energy consumption after active energy efficiency improvements falls below the reference in 7 out of 9 cases. The two cases which have a too high energy consumption are the ones where the window renovation is chosen by its economic feasibility, so the targets can be reached that way. Renovation of the building envelope is thus not obligatory to meet statutory requirements, if renovation is not needed for other reasons, such as indoor thermal comfort, or it is not feasible.

Discussion

The results of this study show that the active energy efficiency improvements are generally economically feasible, and in most cases clearly more feasible than the passive improvements. There is significant savings potential in operating costs induced by executing active energy efficiency improvements. In addition to reducing operating costs, active energy efficiency improvements such as heat pumps may offer considerable emission savings, depending on how the electricity is produced and the primary energy source of local district heating. With an emissions factor for district heating in Lappeenranta of 148 gCo2e/kWh (Lounasheimo et al., 2020), emissions factor of renewable electricity produced by wind power of 10 gCO2e/kWh (Koffi et al., 2017), and SCOP of 1.8, energy-use related emissions could be decreased by 96% just by switching the heat production method from district heating to AWHPs. Obviously, more specific emission reduction calculations would be needed for each building.

The results of the analysis raise one fundamental question: if active energy efficiency improvements are profitable and have a positive impact on the climate, as found in this paper, why are they not widely implemented? The most significant barriers to implementation of energy efficiency technologies are resistance to new technology, lack of knowledge and high initial investment costs (Darko et al., 2017). In context of cold climates, a lack of clear retrofitting regulations is also mentioned (Felius et al., 2020). Other challenges in executing energy efficiency investments are overestimating the costs, cost–benefit mismatch caused by information asymmetry and lack of attention to energy costs, among others (Zhang et al., 2018).

An option to overcome one of the main barriers to adoption of more efficient technology, namely the lack of finance to invest in active energy efficiency improvements and buildings maintenance in general, would be to sell publicly owned properties to private investors, who would execute the energy investments as well as the other renovation. In this Public–Private Partnership (PPP), the public sector would have a sell-and-lease-back agreement with the investor. Thus, the public sector gains access to renovated and energy efficient properties with lower operational costs, and the generation of CO2 emissions would decrease. The investor receives rental income, and the improved energy efficiency reduces the operational costs of the building. Litjens et al. (Litjens et al., 2018) studied combined systems of a ground source heat pump (GSHP), solar photovoltaic (PV) system, and battery storage and pointed out that there is a clear correspondence between the attractiveness of the investment and the total avoided CO2 emissions during a building’s lifecycle. This finding suggests that the level of risk of the investment decreases as the risk to the climate decreases.

The economic analyses of the energy efficiency improvements suggest that the Public–Private Partnership could be a potential way of securing funding of energy efficiency improvements and building renovation and accelerate reduction in CO2 emissions because the more efficient technologies bring remarkable savings in operating costs. If the lease-back period is 20 years and the simple payback time of the energy efficiency investments is about 7 years, the investor will profit from the investments already during the lease-back period. From the municipality’s point of view, the investments are profitable as they would decrease the operational costs of the building. While the energy efficiency improvements are economically feasible, more private parties could be interested in them as investments, which could lead to a situation where PPP methods were more attractive to the public party than they are in the present.

Considering other barriers, namely lack of knowledge, information, and motivation, the solutions to overcome the challenges are rather social and political in nature. More detailed research should be carried out in order to find the ways to overcome the barriers of adopting energy efficient technologies in building sector.

Conclusion

This work analyzed the profitability of investment in energy efficiency improvement in Finnish climate condition for 12 case buildings based on real consumption data.

The key findings were:

  • Considerable savings in operating costs can be achieved by implementation of energy efficiency improvements to existing buildings, confirming the first hypothesis. The total savings potential when the most appropriate technologies for each building are implemented is on average 35% of yearly energy-related operating costs.

  • Active energy efficiency improvements are more profitable than passive energy efficiency improvements, confirming the second hypothesis. Thus, for example, the replacement of existing windows is not competitive investment in the viewpoint of savings although it might be needed for other reasons. Also, the legislation of energy efficiency in renovation of buildings should focus more on active energy efficiency improvements rather than on passive improvements.

  • The assessment of specific building types in relation to the potential for savings in the operating costs shows that there is significant potential for savings in all buildings studied, regardless of the type or the age of the building.

  • The observed savings in operating costs yielded to such returns on the investment that it should make the Public–Private Partnership an attractive option for all parties involved. Public–Private Partnership would allow municipalities and cities with large property volumes to decrease building operating costs and accelerate reduction in greenhouse gas emissions, showing the way to a sustainable future.