Integrating the Assessment of Environmental Costs and the Non-Energy Benefits of Energy Efficiency into an Energy Demand Analysis of the Tertiary Sector

Abstract

Energy system models or energy demand analysis, such as top-down and bottom-up models, provide energy consumption data in energy end-uses, energy carriers, and subsectors. A technical energy efficiency potential can be determined by applying the best available technology (BAT) values. This paper aims to take the consideration of the energy efficiency potential in an energy system model to a new level by including environmental and social aspects. Using the example of the tertiary sector, it is shown how to incorporate a quantification methodology for non-energy benefits (NEBs) together with the (avoided) environmental costs into the energy efficiency potential of an energy demand analysis. This leads to an overall environmental assessment of the tertiary sector and shows how integrating avoided environmental costs and NEBs increases the economic efficiency, cost-effectiveness, and profitability of energy efficiency measures. Assuming a price of 29.04 EUR-ct₂₀₂₀/kWh for the (avoided) environmental costs and a ratio of the total net savings to the net energy savings of 2.3 for the NEBs, all considered energy efficiency measures are economic. This paper shows that including environmental costs and considering NEBs could be important policy instruments.

1. Introduction

1.1. Background

With the average global temperature exceeding the Paris Agreement’s limit of 1.5 °C above the pre-industrial average, 2024 was the warmest year on record according to the EU’s Copernicus Climate Change Service [1]. One implication of this is that greater efforts will be required to implement more ambitious climate change plans and to strengthen the three sustainability strategies of sufficiency, consistency, and efficiency (especially energy efficiency) [2]. In addition, there are rising costs due to environmental damage, and investments in climate change adaptation can no longer be postponed [3,4,5]. Costs could be damage to ecosystems and nature or the impacts of pollutant emissions. The costs of damage to the environment are among the rarely considered external costs of energy consumption [6]. Externalities are costs not included in the market price because their price has not been determined [6,7]. External costs should be taken into account in macroeconomic models and may even be necessary to obtain a complete picture of the costs incurred [8]. Therefore, the concept of environmental costs is one way for policymakers to allocate the costs of environmental damage and climate adaptation measures, in whole or in part, to those responsible [9,10]. Besides the additional costs of externalities, there are also a variety of benefits besides the energetic view for companies, society, and the environment resulting from, e.g., a reduction in energy consumption due to increased energy efficiency. These are referred to as non-energy benefits (NEBs). In addition to lowering energy costs and reducing CO₂ emissions, NEBs include reduced waste, lower pollutant emissions, increased production, or an improved working environment and public image. These NEBs should also be included in the external costs or benefits to obtain comprehensive results. However, economic calculations do not traditionally include external costs, such as environmental costs or NEBs, in the investment process [6]. Similarly, policies do not yet set or specify any conditions for their inclusion [11]. In addition, external costs need to be internalized if they are to be applied to the market. The difficulty here is that external costs can only be determined to a very limited extent due to data availability. Previous studies have indeed analyzed NEBs and developed a framework for including them in investment processes [12]. So far, however, environmental costs, non-energy benefits, and energy efficiency potentials have been considered separately in the literature, and a combination of all three is still rare.

When calculating the energy efficiency potential in an energy demand model, including the environmental costs or the (avoided) environmental costs and NEBs, the cost-effectiveness and persuasiveness of energy efficiency measures are improved. Therefore, this paper aims to develop a methodology to integrate (avoided) environmental costs and non-energy benefits when calculating the energy efficiency potentials in an energy demand model using the example of the tertiary sector.

1.2. Research Gap

This paper addresses the following research questions: Why are NEBs and environmental costs not included in conventional energy demand modeling yet, how can (avoided) environmental and non-energy benefits be included in the energy efficiency potential, and how do they influence it? After literature research on how environmental costs and non-energy benefits can be quantified and monetized, a generally applicable methodology was developed to incorporate them when calculating the energy efficiency potential. This is demonstrated using the example of energy efficiency measures in an energy demand model of the tertiary sector. In addition, sensitivity analyses were conducted.

2. Literature Review

The different strands of the literature on energy efficiency potentials of energy demand models, environmental costs, and NEBs have so far only been analyzed individually.

2.1. Techno-Economic Assessments of Energy Efficiency Measures

Traditional energy system models describe forecast or backcast energy demand or supply [13,14]. In addition, they can show energy efficiency potentials. Fleiter et al. [15] differentiated four categories of energy efficiency potential. They compared an economic potential scenario, a cost-effective scenario, and a technical potential scenario with a frozen efficiency and business-as-usual scenario. To integrate energy efficiency measures into investment processes, the specific cost of conserved energy (SCCE) can be displayed in a graph [16]. If the curve is below the y-axis, this indicates that the efficiency measure is economical.

The approach in this paper is based on the integrated bottom-up and top-down energy demand model of the tertiary sector in Germany described in Arnold-Keifer et al. [17]. A further evaluation of the modeling of technical energy efficiency potentials was conducted in Arnold-Keifer et al. [18]. This paper showed the dependency of energy efficiency potential calculations on the input data and methodology chosen [18].

2.2. Quantification of Environmental Costs

Environmental costs refer to any expenses associated with the environment that may arise from factors such as the depletion of natural resources that businesses rely on, effects on human health and ecosystems, or costs incurred to comply with environmental regulations [19]. As environmental costs are incurred by society but have not yet been included in the market price, they are counted as external costs [6,20,21]. When it comes to comparing potential savings due to energy efficiency measures, these costs are also referred to as avoided environmental costs [22]. In general, the calculation methods of environmental costs can be divided into non-monetary and monetary calculation methods [23,24,25]. This paper focuses on the monetary approach, and

Table 1. Monetization methods for environmental costs.

Monetization Method for Environmental Costs		Methodology	Challenge	Reference
Damage costs (costs for environmental and health damage and damage reduction and prevention)	Market price approach	Damage costs valued at market prices and economic value	No market for environment	[9,26,27,29]
	Stated or revealed preference	Connection of a non-market good to a replacement good	Based on choice experiments and information level of participants	[22,29]
	Societies willingness-to-accept (WTA)	Compensation cost society is willing to accept for compensating the damage	Based on survey (different financial and local situation of respondent)	[30,31]
	Societies willingness-to-pay (WTP)	The amount of money society is willing to spend to prevent environmental damage	Based on survey (different financial and local situation of respondent)	[22,31]
Abatement/Avoidance costs	Standard price approach	Costs of measures to achieve certain political reduction targets	Political definition	[26]
Replacement costs		Compensation costs for negative impacts	Multiple impacts on characterization	[22]

shows an overview of the monetization methods used. The damage cost approach can be further subdivided into the market price approach, the revealed or stated preference approach, the willingness-to-pay (WTP), and the willingness-to-accept (WTA) approaches [24,26,27,28,29].

In this paper, the monetary value is derived from the damage costs for climate change and air pollution. Despite the problem that there is no market for environmental goods, the damage cost approach for climate change is widespread, as can be seen in the Interagency Working Group on Social Cost of Greenhouse Gases [32], EPA [33], Bünger and Matthey [9], the Intergovernmental Panel on Climate Change [34], and Rennert et al. [35] (

Table 2. Monetary values for environmental damage costs of climate change.

Literature	Discount Rate	Costs in EUR2020/ t CO2-eq
Interagency Working Group on Social Cost of Greenhouse Gases [32]	2.5%	67
	3.0%	45
	5.0%	12
EPA [33]	1.5%	298
	2.0%	167
	2.5%	105
Intergovernmental Panel on Climate Change [34]	0.0%	237
	1.0%	160
	3.0%	29
Bünger and Matthey [9]	0.0%	680
	1.0%	195
Rennert et al. [35]	1.5%	270
	2.0%	162
	2.5%	103
	3.0%	70

).

Table 3. Monetary values for environmental damage costs of air pollution [9,37].

Air Pollutant	Specific Emissions in g/kWh	Total Costs in EUR2020/t
Dust	0.01	23,300
PM10	0.009	23,300
NO2	0.374	15,340
SO2	0.196	14,930
NMVOC	0.015	2300
Total costs in EUR-ct2020/kWh	-	0.911

shows the composition of the monetary values for environmental damage costs of air pollution [9,36,37]. The total costs are the sum of the costs of each pollutant multiplied by the specific emissions.

The calculation of the total environmental costs can be seen in Table 4. The first column of the damage costs of climate change with the corresponding discount rate is initially converted to standardized EUR-ct₂₀₂₀/kWh (as in Table 2) and multiplied by the specific CO₂ equivalent emissions of the German electricity mix to obtain the value in the third column. For further investigations, the values of Bünger and Matthey [9] were selected from Table 2 to cover the highest value, as well as the values of Rennert et al. [35], as these cover the full range of discount rates. For the purposes of comparison in this paper, the monetary value assigned to the environmental damage costs of climate change due to electricity generation is 4.44 EUR-ct₂₀₂₀/kWh with a discount rate of 2.5% and is taken from Rennert et al. [35]. To take further aspects of environmental costs into account, the environmental damage costs of air pollutant emissions of 0.911 EUR-ct₂₀₂₀/kWh (based on Table 3) are added based on the Federal Environment Agency’s differentiated cost rates for power plants (no discount rate given) [9].

In addition, to avoid the double-counting of climate damage costs, an annual value of the auction price weighted by the volume auctioned under the EU ETS in 2020 of 24.37 EUR₂₀₂₀/tCO₂ is subtracted [38]. The values in italics represent alternatives for sensitivity analysis. These values are then added to the energy source price of 23.03 EUR₂₀₂₀-ct/kWh to internalize the external costs [36,39].

Table 4. Environmental damage costs [31,33].

Costs of Climate Change [9,35]	Discount Rate	Costs of Climate Change [EUR-ct2020/kWh]	EU Emissions Trading System (ETS) Auction Price in 2020 (24.37 EUR2020/tCO2 [38]) [EUR-ct2020/kWh]	Damage Costs of Air Pollutants [EUR-ct2020/kWh] [9,36]	Environmental Damage Costs [EUR-ct2020/kWh]
EUR2020/CO2,eq: 680	0.0%	29.17 1,3	−1.045 3	0.911 4	29.04 1
USD2020/CO2,eq: 308	1.5%	11.59 1,2,3	−1.045 3	0.911 4	11.46 1,2
USD2020/CO2,eq: 185	2.0%	6.96 1,2,3	−1.045 3	0.911 4	6.83 1,2
USD2020/CO2,eq: 118	2.5%	4.44 2,3	−1.045 3	0.911 4	4.31 1,2

¹ used for the sensitivity analysis. ² converted from USD₂₀₂₀ to EUR₂₀₂₀ using the exchange rate 0.877 [40]. ³ Specific CO₂ equivalent emissions are used for the German electricity mix (429 gCO₂-eq/kWh in 2020) that take into account all upstream emission chains [41]. ⁴ converted from EUR₂₀₂₂ to EUR₂₀₂₀ [42].

Table 4. Environmental damage costs [31,33].

Costs of Climate Change [9,35]	Discount Rate	Costs of Climate Change [EUR-ct2020/kWh]	EU Emissions Trading System (ETS) Auction Price in 2020 (24.37 EUR2020/tCO2 [38]) [EUR-ct2020/kWh]	Damage Costs of Air Pollutants [EUR-ct2020/kWh] [9,36]	Environmental Damage Costs [EUR-ct2020/kWh]
EUR2020/CO2,eq: 680	0.0%	29.17 1,3	−1.045 3	0.911 4	29.04 1
USD2020/CO2,eq: 308	1.5%	11.59 1,2,3	−1.045 3	0.911 4	11.46 1,2
USD2020/CO2,eq: 185	2.0%	6.96 1,2,3	−1.045 3	0.911 4	6.83 1,2
USD2020/CO2,eq: 118	2.5%	4.44 2,3	−1.045 3	0.911 4	4.31 1,2

¹ used for the sensitivity analysis. ² converted from USD₂₀₂₀ to EUR₂₀₂₀ using the exchange rate 0.877 [40]. ³ Specific CO₂ equivalent emissions are used for the German electricity mix (429 gCO₂-eq/kWh in 2020) that take into account all upstream emission chains [41]. ⁴ converted from EUR₂₀₂₂ to EUR₂₀₂₀ [42].

2.3. Quantification of Non-Energy Benefits

Improving energy efficiency often brings benefits beyond energy-related aspects. These are known as ‘non-energy benefits (NEBs)’ or ‘multiple benefits’ [43,44,45]. Non-energy benefits of energy efficiency describe the effects of energy efficiency measures in addition to reducing energy consumption and energy costs, including, for example, the generation of additional value added or positive health effects [44,45,46,47]. Therefore, this concept captures the extent to which energy efficiency measures have other positive side-effects. Fleiter et al. [48] proposed a classification scheme for energy efficiency measures including NEBs. Nehler and Rasmussen [12] developed a framework to include NEBs in investment processes and conducted a survey among companies on how they consider NEBs. The results showed that almost all the companies observed NEBs in maintenance, operation, and the work environment, but only a few know how to monetize them [12]. The consideration of NEBs provides a more holistic picture of the impact of an energy efficiency measure, making it easier to analyze its sustainability. The main established categories of NEBs are shown in

Table 5. Categories and indicators of non-energy benefits [44,45,49,50,51].

Category	Impact	Company-Level	Society
Environmental	Air pollution	X	X *
	Greenhouse gas emissions (GHG)	X	X *
	Reduction of waste, hazardous waste, and water losses	X	X
	Resource management	X	X
Social	Health	X	X
	Well-being	X
	Improved air quality and temperature control	X
	Poverty reduction		X
Economic	Increased production	X
	Product quality	X
	Reduced maintenance costs/extended service life	X
	Reduced costs for compliance with environmental regulations	X
	Innovation and competitiveness	X
	GDP		X
	Employment		X
	Energy prices		X
	Energy supply and security		X
	Energy savings	X	X

* Part of environmental cost methodology.

. NEBs can have an impact both at the company level and at the societal level. Less air pollution, for example, can lead to less pollution and better air quality locally at the company level, just as it leads to better air quality at the societal level. (Avoided) environmental costs can also be quantified as an NEB in a macroeconomic perspective, but clear quantification methods exist (see Section 2.2). The quantification of NEBs in this paper is limited to studies at the company level. In this way, double counting can be avoided.

The literature on NEBs can be divided into quantifying and non-quantifying approaches on the company level; only the quantifying ones are discussed in the following. Pye and McKane [51] used a measure-level approach for quantification, where a financial analysis was carried out for each project. Worrell et al. [49] have analyzed more than 70 case studies from the industry and included them in conservation supply curves (CSC). Zhang et al. [52] evaluated NEBs for climate change and air pollution in China’s cement industry. Combining NEBs in the residential sector, like noise reduction, better room air quality, and comfort, with energy efficiency investments was analyzed in Jakob [53]. Thema et al. [54] presented a model to assess the impact on health, ecosystems, and resources of reducing air pollution, and they included social impacts. Another option was introduced by Reuter et al. [45], who applied a set of indicators using specific calculation methods. Wagner et al. [55] developed a three-stage model methodology based on delimitation, assessment, and evaluation. However, all these different approaches to quantify NEBs show some limitations, such as a lack of transparency in the application of the method, limited applicability or transferability at the company or sector level, or the need for detailed data in particular to obtain an estimate of the influence of NEBs.

In our paper, Arnold-Keifer et al. [56], we describe a practicable way to quantify NEBs based on available data and the calculation of the net present value of an investment, which is shown as follows (Equation (1)):

$$\mathrm{N}\mathrm{P}\mathrm{V}=-{\mathrm{C}}_0+\sum _{\mathrm{t}=1}^{\mathrm{n}}{\displaystyle \frac{{\mathrm{C}}_{\mathrm{t}}}{{(1+\mathrm{r})}^{\mathrm{t}}}}$$

(1)

where

NPV: net present value;

C₀: initial investment;

C_t: net cashflow in time t;

r: discount rate;

n: duration of investments.

The cashflow C_t is assumed to be constant (C_t = C) if it affects the entire lifetime. A reduction of C_t represents the energy savings. The calculation of the annuity of the present value is shown in the following Equation (2):

$$\mathrm{P}\mathrm{V}=\sum _{\mathrm{t}=1}^{\mathrm{n}}{\displaystyle \frac{1}{{(1+\mathrm{r})}^{\mathrm{t}}}}={\displaystyle \frac{1-{(1+\mathrm{r})}^{-\mathrm{n}}}{\mathrm{r}}}$$

(2)

where

PV: present value of annuity.

This results in the following Equation (3):

$$\mathrm{C}={\displaystyle \frac{\left(\mathrm{N}\mathrm{P}\mathrm{V}+{\mathrm{C}}_0\right)·\mathrm{r}}{1-{(1+\mathrm{r})}^{-\mathrm{n}}}}$$

(3)

The annual total net savings S_total include the annual net energy savings S_ener and the annual monetized NEBs S_NEB:

$${\mathrm{S}}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}={\mathrm{S}}_{\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}}+{\mathrm{S}}_{\mathrm{N}\mathrm{E}\mathrm{B}}$$

(4)

where

S_total: annual total net savings;

S_ener: annual net energy savings;

S_NEB: annual monetized NEBs.

With Equations (3) and (4), S_ener can be calculated with the following Equation (5):

$${\mathrm{S}}_{\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}}={\displaystyle \frac{\left({\mathrm{N}\mathrm{P}\mathrm{V}}_{\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}}+{\mathrm{C}}_0\right)·\mathrm{r}}{1-{(1+\mathrm{r})}^{-\mathrm{n}}}}$$

(5)

where

NPV_ener: net present value of the investments taking into account the pure energy savings.

Likewise, S_total can be calculated with the following Equation (6):

$${\mathrm{S}}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}={\displaystyle \frac{\left({\mathrm{N}\mathrm{P}\mathrm{V}}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}+{\mathrm{C}}_0\right)·\mathrm{r}}{1-{(1+\mathrm{r})}^{-\mathrm{n}}}}$$

(6)

where

NPV_total: net present value of the investments, taking into account the total investments.

The ratio between the total net savings S_total (Equation (6)) and the net energy savings S_ener (Equation (5)) shows the magnitude of the effect of NEBs on energy savings.

$$\mathsf{\theta }={\displaystyle \frac{S_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}}{{\mathrm{S}}_{\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}}}}$$

(7)

where

θ: ratio of the total net savings S_total to net energy savings S_ener.

A value greater than one is an indication that there are significant NEBs in addition to energy savings [56]. Based on 16 case studies from [51,57,58], a median value of 1.6, an arithmetic mean of 2.3, and a 25% quantile of 1.3 were calculated for the ratio θ of total net savings to net energy savings [56]. This paper also shows that the annual net energy savings can be multiplied by the ratio to estimate the savings that could result from incorporating NEBs.

2.4. The Need to Integrate Environmental Costs and NEBs When Calculating the Energy Efficiency Potential

Previous studies have analyzed environmental costs, NEBs, or energy efficiency potentials, but the combination of these three aspects is still rare. At present, the individual methods tend to be used separately, as can be seen in

Table 6. Different strands of the literature on energy demand modeling with energy efficiency potential, non-energy benefits, and environmental costs.

	Traditional Energy-Demand Modeling and Energy Efficiency Potential	Non-Energy Benefits	Environmental Costs
Perspective	Individual/microeconomic/macroeconomic	Company level, extended to national level by IEA [44]	Societal/environmental/sustainability perspective
Focus	Energy demand and possible efficiency potential	Benefits of energy efficiency measures beyond the energy perspective [49]	Any costs related to the damage of the environment [19]
Quantification	Resulting energy demand: Frozen efficiency, economic potential, technical potential [48]	Quantifying and non-quantifying approaches	Few studies, ‘virtual’ costs for companies
Challenge	Not a complete picture of environmental and sociological influences	Poor data quality and availability, especially for the tertiary sector	Few studies on quantification; not implemented by politicians

, as they have emerged from completely different directions and have, therefore, been applied separately up until now. Environmental costs tend to come from the sustainability perspective, while NEBs were initially seen as business benefits but are now also considered at the national level, and energy models have traditionally been characterized by the energy demand perspective. However, energy efficiency can be taken as the connecting link here.

This paper aims to combine the existing approaches and integrate them into an energy system model based on the example of the tertiary sector. NEBs and energy efficiency are already closely linked, as NEBs result from energy efficiency measures, and the avoided environmental costs can also be seen as NEBs in economic terms. Moreover, Mzavanadze [59] combined environmental costs, or parts thereof, with energy efficiency measures but did not give specific values or a monetization formula. Sovacool et al. [60] aggregated data from studies that analyze the external costs of electricity supply in a global perspective, without looking more closely at the actual investment process. With the method described in this paper, all existing frameworks are to be extended by the respective others, including the energy efficiency potential, the concept of environmental costs, as well as the NEBs (see Figure 1).

Since it is very important to consider NEBs and environmental costs in investment processes in order to obtain a complete view of all costs of the process, the three sides of the energy efficiency medal (energy efficiency potential, NEBs, and environmental costs) should be included in an energy model for an overall societal and broadly conceived approach. This paper aims to close this gap and integrate environmental and social aspects by developing and verifying a generally applicable methodology to integrate (avoided) environmental costs and NEBs when calculating the energy efficiency potential of measures. The combination and quantification of the different approaches of energy efficiency potential, environmental costs, and NEBs are important potential policy measures. All costs and benefits of energy efficiency measures are presented transparently in a single calculation and can be used for all applications of electricity consumption. This is an important and novel contribution and can be a supplement or substitute to existing instruments in market-based energy policies. Other existing market-based policy instruments are obligations like white certificates that declare a certain decrease in energy consumption, auctions like emission taxes or emission trading like the European Trading System (ETS) regulating energy-intensive industries, fossil power generation plants, European air and maritime transport, and, from 2027 on, fuel combustion in buildings, road transport, and small industry sectors [61,62,63,64,65]. Another possibility is an efficiency-oriented energy-consuming right trading system like energy quota trading [66,67].

2.5. Tertiary Sector

If environmental costs were applied by politicians, this would have a significant effect on the tertiary sector and is, therefore, used as an example of an energy demand model. The tertiary or service sector is very heterogeneous and accounted for 14.7% of Germany’s total energy consumption in 2019 [68]. Tsemekidi Tzeiranaki et al. [69] analyzed the tertiary sector in the period from 2000 to 2019 in the EU27 plus UK and noted a continuous increase in final energy consumption. At the EU regulatory level, the tertiary sector is closely linked to the building sector. Therefore, the directive from Article 5 of the EED must also be enforced here, according to which the total final energy consumption of all public facilities has to be reduced by a minimum of 1.9% annually compared to the baseline year 2021 [70]. The findings of Tsemekidi Tzeiranaki et al. [69] suggest that a more ambitious political goal is required, with energy efficiency playing a bigger role, and that the different energy end-uses in the tertiary sector must be examined in more detail. The tertiary sector must also introduce energy-efficient equipment and structures. Policy measures supporting increased energy efficiency could include information and communication, compensation, regulation, incentives or legislation, and financial or tax-related measures [69,71]. Some studies also describe the impact of NEBs in the tertiary sector. For example, Neusel and Hirzel [50] examined cold-supply food chains and found that, in addition to cost savings through energy efficiency, lower production costs also occurred. Highly-efficient buildings provide health and productivity benefits, as well as better indoor air quality and thermal comfort [54,72].

3. Methodology

3.1. An Integrated Methodology to Quantify Environmental Costs and Non-Energy Benefits

For a more holistic perspective, environmental costs and NEBs should be part of the full cost calculation of energy efficiency measures within an investment process. Our approach is an attempt to combine two strands of the literature and to show that this combination is useful, necessary, and feasible. Therefore, we include environmental costs and NEBs in the specific costs of conserved energy (SCCE) by adding the environmental damage costs of electricity generation onto the energy price and considering NEBs. This results in the annual energy cost savings that take external costs into account. According to our paper, Arnold-Keifer et al. [36], the specific costs of conserved energy (SCCE), including (avoided) environmental costs, are calculated as shown in Equation (8). NEBs with a positive effect on energy cost savings and the avoided environmental costs as ‘virtual’ costs for the company have an overall positive effect on the net present value, shown as follows:

$$\begin{array}{c}SCCE={\displaystyle \frac{net present value}{present value of energy savings}}\hfill \\ ={\displaystyle \frac{I-\sum _{n=1}^N\frac{\left(E_n\cdot P_n \pm O\&M_n \pm {EC}_{ne}\right)}{{\left(1 + d\right)}^n}}{{\sum }_{n=1}^N\frac{E_n}{{\left(1 + d\right)}^n}}}\hfill \\ ={\displaystyle \frac{I-\sum _{n=1}^N\frac{{\left(E·\left(P + {EC}_{kWh}\right) \pm ∆O\&M \pm {EC}_{ne}\right)}_n}{{\left(1 + d\right)}^n}}{{\sum }_{n=1}^N\frac{E_n}{{\left(1 + d\right)}^n}}}\hfill \\ \equiv {\displaystyle \frac{(I-\sum _{n=1}^N\frac{{\left(E·\left(P + {EC}_{kWh}\right) \pm ∆O\&M \pm {EC}_{ne}\right)}_n}{{\left(1 + d\right)}^n})·AF}{E_n}}\hfill \end{array}$$

(8)

where

$$AF={\displaystyle \frac{d}{1-{(1+d)}^{-n}}}$$

(9)

and

SCCE: specific cost of conserved energy [EUR/kWh];

I: investment [EUR];

d: discount rate [1/yr];

n: lifetime [yr];

E: energy saving potential [kWh/yr];

P: energy price [EUR/kWh];

EC_kWh: environmental costs of electricity generation per kWh [EUR/kWh];

$O\&M:$ operation and maintenance costs [EUR/yr];

EC_ne: non-energy benefits [EUR/yr];

AF: annuity factor.

A number of assumptions are made in this model. The energy saving potential stays the same per year over the lifetime, as does the price of the energy source. The investments are considered over the lifetime.

The theoretical consideration of the formula already shows the influence of the discount rate on the net present value. If the discount rate is increased, then the net present value decreases. If the investments were zero, then the SCCE would be the sum of the energy source price plus the change due to the inclusion of the avoided environmental costs and NEBs in relation to the energy savings.

This paper aims to show how the method of integrating environmental costs and non-energy benefits is applied. It should be noted that the data used are limited and are only intended to illustrate the methodology. Maintenance costs, for example, are not considered in the measures analyzed in this paper.

3.2. Data Sources and Sensitivity

The technical energy efficiency potential was obtained from an integrated bottom-up and top-down energy demand model of the tertiary sector in Germany introduced in our paper Arnold-Keifer et al. [17]. Using the data from a survey of 1451 companies in 2019, the final energy demand can be subdivided by energy carrier, energy end-use, and subsectors, as shown in Figure 2 [17]. The model is structured as follows: At the end-use level, the model determines the final energy demand per device in each company using direct data from the survey. These bottom-up data are adjusted by the top-down energy demand for each company stated in the survey. In the next scaling step, these data are extrapolated to a sub-sectoral level using drivers, including the number of employees, students, and annual guests in accommodation or hospital beds. The last step includes the extrapolation to the energy balance of the tertiary sector [68]. This final step is skipped in this paper, as there is no comparable energy balance for Germany that includes energy efficiency measures. Moreover, there is only an aggregated deviation of 1% between the model and the energy balance. This paper only analyzes the electricity demand and assumes that the energy demand derived from this model corresponds to the state-of-the-art in the tertiary sector. Energy efficiency measures are considered in a second step, and the best-available technique (BAT) values replace the energy consumption values provided by the company in the survey. The difference between the companies’ stated energy consumption and the BAT values reveals the technical energy potential. The best-available technology values for most products are based on the European Commission’s Ecodesign and Efficiency Labeling regulations.

Exceptions were made for compressors and ventilation due to the lack of data from the survey and because it is not possible to use a BAT value here. Instead, an assumed value for the savings per system was taken based on the evaluation of federal funding for energy and resource efficiency in the German economy [73]. There are some more restrictions: Servers were not replaced by BAT values but were kept at a constant level, as the development process is not aimed at increasing efficiency by replacing devices, but at achieving a leap in efficiency through technological progress and new applications. The number and structural background of industrial processes used by the companies are not specified in more detail in the survey and, therefore, could not be replaced with a more efficient product either. The specific energy consumption for providing food in company canteens depends on numerous influencing factors that make it difficult to estimate the investment costs for the companies, so these are also kept constant. There are similar problems with estimating construction costs for investing in cold storage rooms, freezer rooms, indoor swimming pools and saunas, and in the subsectors of agriculture, market gardens, and fisheries. Investment costs could not be calculated for washing machines either, as the survey did not collect any information about the washing machines used, and energy consumption was only estimated based on the weight of laundry washed.

The energy-saving potential is used as a starting value for integrating environmental costs and non-energy benefits. Energy efficient technologies are always defined in comparison to a reference or standard value [74]. Therefore, the investment costs can be seen as the differential costs or replacement investment between a business-as-usual (BAU) purchase and the immediate minimum requirement for eco-design, the least life cycle costs (LLCC) [75]. The sum of the replacement investment costs is calculated per unit multiplied by the number of devices. The number of devices from the survey was extrapolated to estimate the number of devices in Germany as a whole. Other required data, such as discount rates, lifetimes, and energy carrier prices, were derived from the literature or estimated. The values used are shown in

Table 7. Input data for the tertiary sector.

Energy Efficiency Measures In Energy End-Use	Application [References]	Life-Time2 [yr]	Annual Energy Saving Potential [TWh]	Investment (BAU–LLCC) [EUR] Per Total Energy Cost Savings [EUR]	SCCE [EUR-ct2020/kWh]	SCCE Including Environmental Costs [EUR-ct2020/kWh]	SCCE Including Environmental Costs and NEBs [EUR-ct2020/kWh]
Lighting (I)	LED [76]	15	8.69	0.05	−0.220	−0.263	−0.401
Air conditioning (II)	Central air conditioning [77]	15	1.68	0.06	−0.217	−0.260	−0.398
	Mobile air conditioning [77]	15
	Decentralized split system [77]	15
Mechanical energy (III)	Lift [29,78]	25	0.43	0.29	−0.163	−0.206	−0.344
	Ventilation 1 [73,79]	25
	Compressor 1 [73]	25
ICT (IV)	Desktop Computer [80,81]	5	5.08	0.35	−0.149	−0.192	−0.330
	Laptop [80,81]	5
	Monitor [80,82]	5
	Printer [80,83]	5
	Copy machine [80]	5
	Projector [80]	5
	Cash register 2	5
Process heat (V)	Dishwasher [84,85]	15	0.17	0.67	−0.075	−0.118	−0.257
Process cooling (VI)	Refrigerator [86,87]	15	0.07	2.88	0.432	0.389	0.251
	Freezer [86,87]	15

¹ energy consumption for the savings per system was taken based on the evaluation of federal funding for energy and resource efficiency in the German economy [73]. ² estimation.

. For the calculations, the measures taken to improve the energy efficiency of the applications are grouped according to the respective end-uses.

The energy saving cost curve was recalculated for various scenarios for each electricity end-use to obtain a better understanding of the impact on the curve. In our study, the environmental costs and the discount rate were subjected to variations, as was the ratio of NEBs of total net savings to net energy savings, as shown in

Table 8. Variations of environmental costs and the ratio of NEBs.

	Environmental Costs [ct/kWh]	NEBs [Ratio]	Discount Rate
Basis	-	-	0.025
Variation (1)	4.31	1.6	0.025
Variation (2)	4.31	2.3	0.025
Variation (3)	6.83	1.6	0.020
Variation (4)	6.83	2.3	0.020
Variation (5)	29.04	2.3	0.000

.

The first variations aimed to quantify NEBs. Therefore, the initially selected median ratio between the total net savings and the net energy savings of 1.6 was changed to the arithmetic mean of 2.3, as determined in Arnold-Keifer et al. [56]. Further variations examined the influence of environmental costs and changed the initial value of 4.31 EUR-ct₂₀₂₀/kWh and a discount rate of 0.025% to 6.83 EUR-ct₂₀₂₀/kWh with a discount rate of 0.02 and to 29.04 EUR-ct₂₀₂₀/kWh with a discount rate of 0.00%.

In addition, there are other factors that influence the SCCE curve, such as lifetime, discount rate, energy price, and investments. These were examined in the sensitivity analysis.

3.3. Sensitivity Analysis

One sensitivity analysis aimed to analyze the replacement investment costs. Here, each value of the replacement investments was changed in the range between +10% and −10%. This was to investigate possible errors in the literature values and estimates of investment costs. Another sensitivity analysis investigated the influence of the energy price in the range from −50% to +200%. The influence of the discount rate cannot be considered individually, as the environmental costs are always directly linked to it; thus, the influence is already covered by the consideration of the different variations in Section 3.1. A third sensitivity analysis was carried out with the lifetime. Here, the specified lifetimes were changed by +−20%. Regional differences in the specific emission factor of the electricity mix were investigated in a fourth sensitivity analysis. In addition to Germany, Poland, with a very high value within the EU, and Sweden, with the lowest value, were selected, and the specific environmental damage costs calculated are shown in

Table 9. Environmental damage costs based on the costs of climate change of 118 USD₂₀₂₀/CO_2,eq and a discount rate of 2.5% for different countries [35,88,90,91].

Country	Specific Emission Factor of Electricity Mix [gCO2-eq/kWh in 2020]	Costs of Climate Change [EUR-ct2020/kWh]	EU Emissions Trading System (ETS) Auction Price in 2020 (24.37 EUR2020/tCO2 [38]) [EUR-ct2020/kWh]	Damage Costs of Air Pollutants [EUR-ct2020/kWh] [9,36]	Environmental Damage Costs [EUR-ct2020/kWh]
Germany	429	4.44	−1.045	0.911	4.31
Poland	750	7.76	−1.045	0.911	7.63
Sweden	39	0.40	−1.045	0.911	0.27

[88]. The energy price for 2020 was also adjusted here for the countries. For Poland, 14.75 EUR-ct₂₀₂₀/kWh was selected, and for Sweden, 19.14 EUR-ct₂₀₂₀/kWh was selected [89].

4. Results

4.1. Application of the Methodology to Quantify Environmental Costs and Non-Energy Benefits

The technical energy efficiency potential of the German tertiary sector was compared with the energy efficiency potential including (avoided) environmental costs and NEBs. In both cases, the energy efficiency measures were cost-effective in all end-uses, except process cooling. This is indicated by the fact that the SCCE in the Figure 3 curve drops below the y-axis into the negative range. The economic annual energy saving potential is 16.06 TWh (99.5% of the cumulative annual energy saving potential). Nevertheless, environmental costs and NEBs have no effect on the annual energy-saving potential of 16.13 TWh (with 144 TWh electricity consumption of the entire tertiary sector in Germany in 2019).

Compared to the others, measures V and VI (process cooling and process heat) have a minor cumulative annual energy savings potential. However, this is due to the model’s structure and the resulting lower savings potential. Process cooling has a relatively low savings potential for relatively high investment costs. Lighting, on the other hand, has a very high cumulative annual energy savings potential.

Figure 4 shows the results of variations of the environmental costs, the discount rate, and the ratio for NEBs of total net savings to net energy savings in detail.

Measures I–V were cost-effective in all the sensitivity analyses. However, it must be stated that the different lifetimes and the changed discount rate, due to the increased environmental costs, also influenced the SCCE curve. With the highest value of the environmental costs of 29.04 EUR-ct₂₀₂₀/kWh, with a discount rate of 0.00%, and a ratio for NEBs of 2.3, the energy efficiency measures of process cooling (VI) were also deemed economical. The influence of higher NEBs has a greater impact than higher environmental costs.

4.2. Results of the Sensitivity Analysis

Figure 5 shows the results of the sensitivity analyses of the replacement investment costs in detail. A change in replacement investment costs of +−10% had only a minor impact on the cost curve. The cumulative annual energy saving potential remained the same with 16.13 TWh. This shows that, even with possible deviations in the investment costs, a statement can be made on the economic viability of energy efficiency measures.

The sensitivity analyses of the energy price are shown in detail in Figure 6. The price of energy had a major influence on the curve. As the price of energy and the ETS, which is linked to the price of energy in our calculations, are often subject to fluctuations, this should be taken into account.

Figure 7 shows the results of the sensitivity analyses of the lifetime. When changing the lifetime, it is important to note that the x-axis shows the annual energy cost savings, and the assumption here is that these remain the same despite the change in lifetime. The influence of the lifetime is particularly evident with an already long lifetime, as is the case with mechanical energy. The assumed average lifetime of mechanical energy measures is 25 years. If an already low savings potential meets an even shorter lifetime, then the measure becomes even less economical, as they do for process cooling.

In Figure 8, the influences of different specific emission factors for the electricity mix in different European countries can be seen. Sweden was chosen for the lowest European emission factor and Poland for a relatively high one. However, both countries have a lower electricity price compared to Germany, which distorts the analysis and means that Germany still has the highest overall energy price, including (avoided) environmental costs, despite a significantly lower specific emission factor. However, the trend of Sweden shows that the lower the specific emission factor, the less economical the measures are.

5. Discussion

This paper extended the energy efficiency potential by integrating (avoided) environmental costs and non-energy benefits of energy efficiency measures and applied this concept to an energy demand model of the tertiary sector. Thus, this paper contributes to a more detailed understanding of integrating (avoided) environmental costs and NEBs in an energy demand model. The underlying top-down and bottom-up model was extended to include the technical energy-saving potential by applying the best-available technology values. The resulting efficiency potential was used to calculate the avoided environmental costs and resulting monetary value of the NEBs and extended the energy efficiency potential to include environmental and social aspects in a way not yet described in the literature. The results show that integrating environmental costs and NEBs pulls the cost curve towards the negative range of the y-axis, and this may lead to measures becoming economically profitable. Overall, the measures available for the tertiary sector in the integrated bottom-up and top-down model indicate an annual technical energy-saving potential of 16.13 TWh (with 144 TWh electricity consumption of the entire tertiary sector in Germany in 2019). However, the curve remains parallel to the original energy-saving cost curve.

It can be concluded that, when considering (avoided) environmental costs of 29.04 EUR-ct₂₀₂₀/kWh and a relatively high impact of NEBs (ratio of 2.3), measures become economic that were not before. Changing the environmental costs from 4.31 EUR-ct₂₀₂₀/kWh with a discount rate of 0.025 to 6.83 EUR-ct₂₀₂₀/kWh with a discount rate of 0.02 had less influence than increasing the ratio of the NEBs between the total net savings and the net energy savings from 1.6 to 2.3. It should be emphasized that the discount rate was changed as well and also influenced the specific costs of conserved energy. In addition, the different lifetimes of the measures need to be considered. It should also be noted that other current studies assume significantly higher environmental costs, which could shift the curve even further into the economic range [92]. This can also be seen in the sensitivity analysis of the energy price; a change here has a major impact on the curve. With energy prices constantly fluctuating, this is very important. The geographical studies of different specific emission factors for the electricity mix show that, with a target of 100% renewable energies and a specific emission factor of 0 gCO₂-eq/kWh, the use of environmental costs for electricity-based energy efficiency measures would be obsolete.

It must also be pointed out that the measures within the different end-uses are not complete and, therefore, cannot be interpreted as recommendations in the order of their occurrence in the curve. They are merely an attempt to integrate the environmental and social aspects of the energy efficiency potential into an energy demand model. The energy-saving cost curve should always be seen in the context of additional analyses, for example, a Pareto optimization [93,94].

In addition, the NEBs should be scrutinized as to what has been included in the quantification in order to avoid possible double counting with the environmental costs. In this paper, the NEBs were only quantified at company level and the (avoided) environmental costs were from a macroeconomic perspective. If macroeconomic NEBs are also included, then a double consideration must be taken into account in the internalization.

The results show the possibility of extending the energy efficiency potential to include environmental and social aspects. This approach leads to an overall more accurate assessment of energy efficiency measures in the investment process and better reflects their influence on society as a whole and the true costs and benefits of these measures.

However, integrating environmental costs and NEBs needs careful consideration. First, environmental costs need to be introduced by policies. Second, full internalization of these external environmental costs would raise the energy price. Despite the benefits of considering environmental costs and the argument that the damage should be attributed to the polluter, high political hurdles must be overcome in order to introduce higher energy prices. Policymakers often find it difficult to deal with problems that lead to higher costs that cannot be solved in the short term within a legislative period [95]. Public resistance could also be high if policies do not include measures for the socially disadvantaged who are negatively affected by higher energy prices and energy poverty [62]. In addition, companies would argue that they might have higher compliance costs. Third, the different possibilities and the lack of sufficient data to calculate environmental costs and NEBs must be taken into account, as must the simplifying assumptions made during the calculation. These include the assumption that the energy price, the German electricity mix (429 gCO₂-eq/kWh in 2020), the environmental costs, and the benefits of the energy efficiency measures will remain constant for the next few years. In addition, the EU ETS is subject to fluctuations, which is subtracted from the energy price in this paper. The specific costs of conserved energy are based on an average value for the energy price, which is why no dynamic energy prices can be displayed. However, an estimate of this can be taken from the results of the sensitivity analysis. In future research, the entire model could be extended to include dynamic price scenarios.

For more precise results, detailed data from systematic surveys and studies should be collected to improve the reliability of future studies. More detailed data should be provided for calculating and capturing the NEBs of energy efficiency measures in detail, especially NEB data from the tertiary sector. The methodology could also be extended to the consumption of fossil fuels. Future research could also include and analyze sufficiency measures. This has not yet been taken into account due to the focus on energy efficiency measures and the limited availability of data on possible energy savings from sufficiency measures in the tertiary sector. This would mean that new measures in the area of behavioral changes would have to be added and would require the ability to measure the cost and saving potential of sufficiency measures. These measures could then be included in the calculations in a similar way to the efficiency measures now.

6. Conclusions and Implications for Policies and Modelling Practice

The tertiary sector would be affected by any policy that incorporated the environmental costs associated with electricity use. Energy efficiency measures result in higher avoided environmental costs. Non-energy benefits paint a better picture of the positive influence of energy efficiency measures beyond purely energy-related considerations. The avoided environmental costs and NEBs should therefore be included in the investment process to obtain a more accurate environmental assessment and to truly reflect all the costs and benefits of the investment process and the effects of energy efficiency measures on society. Taking into account the (avoided) environmental costs of 29.04 EUR-ct₂₀₂₀/kWh and the relatively high impact of NEBs (ratio of 2.3 between the total net savings to the net energy savings), all considered energy efficiency measures are economic. This paper shows one way to expand the energy efficiency potential of energy demand models to include environmental and social aspects based on an integrated bottom-up and top-down model of the German tertiary sector.

Our approach can aid politicians and decision-makers in identifying and implementing relevant efficiency measures. The investigations in this paper address the microeconomic level with the energy-related consideration and the NEBs, as well as a macroeconomic perspective with the environmental costs and the NEBs. Policy interventions shown in this paper can improve the economic attractiveness of energy efficiency measures. Even if the measures are uneconomical from a microeconomic level, the consideration of NEBs and environmental costs can justify an allocation of constrained investments that might otherwise not have been economically feasible. The additional revenue from the avoided environmental costs could be used for climate adaptation or climate protection measures. Integrating the (avoided) environmental costs and NEBs could also be transferred to other models in different energy-using sectors. A prerequisite for the transferability to other sectors is a specific data basis for the sector, such as the survey in the tertiary sector in this paper. Furthermore, basic data for energy efficient applications must be available. The concept of NEBs and environmental costs developed in this paper could then be transferred directly.