Energy Demand Modeling for the Transition of a Coal-Dependent City to a Low-Carbon City: The Case of Ulaanbaatar City

Abstract

Cities have committed to reducing greenhouse gas emissions and promoting renewable energy. However, many cities continue to rely on fossil fuels, while renewable energy sources are not used or are unable to meet the demand that fossil fuels provide. Depending on the geographic location, climate, and resources, cities must find their own path to energy sustainability. The city of Ulaanbaatar is one of the coal-dependent cities, its electricity and heat consumption mainly coming from coal. In this study, the future final energy demand of a coal-dependent city is identified and analyzed to make it a low-carbon city. Long-term energy demand projections for Ulaanbaatar to 2050 are conducted using the model for analysis of energy demand (MAED) model. Four scenarios are developed based on the existing local and national policies in the socio-economic and energy sectors, as well as more ambitious policy and technology measures recommended by various studies in the MAED_D model. The final energy demand is calculated to be 548, 460, 334, and 264 PJ in 2050 for BAU, REF, NDC, and RM scenarios, respectively, compared to 135 PJ in 2020. The results show that the high penetration of electricity and renewable energy, energy efficiency measures, and energy intensity reduction in all sectors can significantly reduce the future energy demand and help the transition towards a low-carbon city.

1. Introduction

Cities are key players in the global energy transition and play an important role in the global energy demand and socioeconomic development of countries [1]. Global agendas for sustainable development and climate change emphasize the contribution of cities to the decarbonization of the energy system. Local governments play an important role in achieving the Sustainable Development Goal 7 (SDG 7) targets for energy access, renewable energy, and energy efficiency through a variety of approaches, including setting sustainable energy targets and developing energy master plans that integrate urban energy systems [2]. The Paris Agreement also supports cities in defining their own climate action [3]. Cities only have the potential to achieve the net zero target if they comprehensively decarbonize all sectors, resources, and system transformation. Therefore, cities and their climate actions are critical to achieving a sustainable energy transition and climate change mitigation action at the national to global levels [4]. Cities are increasingly committed to reducing greenhouse gas (GHG) emissions and promoting renewable energy. However, their capacity, planning, and implementation of energy transition policies are still far from the target [1,2,4]. Many cities still rely on fossil fuels, while renewable energy sources are not used or are unable to meet the demand that fossil fuels generate [1]. Depending on the geography, climate, and resources, cities must determine their own energy sustainability. In many cities, it is almost impossible to function without fossil fuels [5].

The city of Ulaanbaatar is one of the most coal-dependent cities in developing countries. It obtains its electricity and heat consumption from coal [6]. Due to the consumption of lignite and bituminous coal, Ulaanbaatar suffers from severe air pollution during the long and cold winter period [7]. The city of Ulaanbaatar has been growing rapidly for decades due to urbanization and economic growth, which increases the demand for energy and heat [8,9]. In addition to socioeconomic growth, policies such as highly subsidized energy tariffs (the price of electricity for households is 25% lower than the cost of production and the price of heating is twice as low as the cost), zero tariff for night electricity in ger areas, subsidies for improved coal, and full exemption of households and businesses from electricity tariffs due to the Coronavirus disease (COVID-19) have led to an increase in energy demand instead of encouraging people to save energy and adopt new technologies [10,11]. Due to the increasing energy demand and economic deficit of utilities, a lack of investment in new energy resources and the delay in the renovation of the existing infrastructure, and heavily subsidized energy tariffs, Ulaanbaatar city is facing a high risk and serious problems of energy shortage and energy security. Therefore, in order to keep up with the increasing energy demand and ensure energy security, it is essential to consider the long-term energy planning and modeling for a sustainable energy transition.

Long-term energy planning helps countries achieve their sustainable development goals. Accurate forecasting results enable decision makers to make an optimal decision to meet future needs [12]. Energy management means the proper use of available resources for sustainable development [13]. Energy demand forecasting is an important tool for policy development and energy management. In many developing countries, this process is challenging due to the lack of data, knowledge capacity, and modeling tools. Energy demand analysis in developing countries differs from that in developed economies in terms of the data availability, energy system, and emerging demand [14]. Therefore, energy demand modeling for developing countries at the city level is carefully selected considering the specific characteristics of Ulaanbaatar city. In the literature, the model for analysis of energy demand (MAED) model is a key model in developing countries because it provides a flexible framework for energy consumption patterns, sectoral disaggregation, and customer use. The MAED model has not yet been used in the Mongolian energy sector.

This paper examines one of the coal-dependent cities, the city of Ulaanbaatar, which has limited institutional and managerial authority over its own electricity and heat supply and a lack of long-term policies consistent with global and national commitments to mitigate climate change and achieve a sustainable energy transition. This study demonstrates and analyzes the potential of the coal-dependent city to transform into a low-carbon city with the lowest possible demand through the implementation of various energy and climate measures and policy scenarios. The main objective of this study is to model long-term energy demand projections and climate change mitigation scenarios for Ulaanbaatar by 2050. The study assesses how the final energy demand will change with the adoption of different policies and which sectors are the largest consumers and how they will evolve in the long term with different policies. The study is the basis and input for the energy optimization modeling of the sustainable energy transition in the city of Ulaanbaatar, Mongolia, which optimizes the low-carbon and least-cost pathways for the transition from the coal-dependent energy system.

This study contributes to the literature on Mongolia’s energy system and the city’s energy transition by (i) systematically examining the energy sector, including the transportation sector and motor fuels, in Ulaanbaatar’s energy system; (ii) applying the generic energy forecasting model of MAED to the energy system of Ulaanbaatar, Mongolia; (iii) analyzing and comparing energy demand forecasting under various scenarios while applying the existing and additional policies for a sustainable energy transition; and (iv) investigating the role of the city and its potential for sustainable energy transition.

Section 1 presents the importance of the role of cities in the sustainable energy transition and the importance of long-term energy planning, as well as the main energy issues in the city of Ulaanbaatar. Section 2 reviews the literature on energy demand forecasting models, the MAED model, and its application in developing countries and cities. In addition, this section presents the previous studies on energy demand forecasting for the city of Ulaanbaatar. Section 3 provides an overview of the energy system of Ulaanbaatar city and its characteristics. The rationale for selecting the MAED model, its methodology, and scenario development are described in Section 4. The model results are discussed in Section 5, and the discussion and suggestions for further studies are presented in Section 6. The policy recommendations are summarized in Section 7.

2. Literature Review

There are numerous energy demand forecasting models that have been developed and evolved since the early 1970s. Bhattacharyya et al. [14] distinguished two broad categories of models: simple and sophisticated approaches. A simple approach is user-friendly in terms of the data and skills; however, it lacks explanatory power, especially in explaining demand drivers and technologies. Sophisticated approaches, econometric approaches, end-use approaches, input–output models, scenario approaches, and hybrid approaches, have been developed to analyze and forecast energy demand through employing more advanced technologies.

The econometric and end-use approaches are widely discussed in terms of their relevance to developing countries and the energy sector, as well as their application in practice. The economic/top-down approach generally focuses on aggregate demands and provides only a limited perspective for energy system analysis. The characteristics of energy systems in developing countries, such as traditional energy transition, urban–rural divide, informal energy supply, differential use of technologies, and structural change issues, are not adequately considered and analyzed in the model [14]. Methodologically, they can be divided into top-down and bottom-up approaches, as well as simulation, optimization, accounting-based, and/or equilibrium models [15]. The end-use/bottom-up approach is based on the engineering economics method, which is used for simulation or optimization purposes and requires disaggregated levels. The end-use approach is well-suited to capture the specific characteristics of developing countries; however, in practice, it is unable to capture informal economies, non-monetary transactions, spatial differences, and different levels of consumption [14].

Forecasting methods can be divided into causal and historical data-based methods. Artificial neural networks and regression models represent the causal methods that consider relationships between various factors and energy consumption to predict energy demands. Time-series, gray prediction, and autoregressive models are historical data-based methods that use historical data to predict future demands [16]. Energy demand forecasts are calculated in different horizons, such as hourly, daily, weekly, monthly, and annually [17].

Bhattacharyya et al. [14] reviewed energy demand models in the context of developing country needs, considering their applicability to traditional energy, inclusion of informal activities, and application to emerging demands. MAED/MEDEE and LEAP models are generic energy forecasting models that can be applied in a variety of contexts. MAED is the most widely used accounting-based, bottom-up, end-use model for medium- to long-term energy forecasting practices. Studies that have used the MAED model at the city level and in developing countries include the city of Belgrade, Serbia, medium-term energy demand projection for the main energy consumers’ households, services, industry, and transport, projected under five different technological alternatives [18]. Nakarmi et al. [19] used MAED to analyze the long-term energy demand for the residential sector in Kathmandu Valley, Nepal. Electricity demand in Cameroon for 20 years was studied using MAED under three economic scenarios [20].

In the draft Energy Master Plan for Ulaanbaatar to 2050, the degree of the long-term decarbonization of UB’s energy system by local and regional renewable energy sources was determined using the KomMod energy modeling tool developed by the Fraunhofer Institute for Solar Energy Systems. Three scenarios were developed based on the available resources of wind, solar, and local fuels in the territory of Ulaanbaatar and the region. Solar, wind, and waste incineration as heating sources were projected to provide 56% of the final energy supply by 2050. The share of imported electricity and CHP ratio for the energy production will be reduced to 2% and 25%, respectively. The final energy supply is increased from 15,553 GWh in 2016 to 16,006 GWh in 2050 [6]. The draft local energy efficiency action plan (LEEAP) for Ulaanbaatar up to 2040 recommends implementing a retrofit program for public buildings, multi-family dwellings, and single-family dwellings in ger areas, to reduce energy consumption levels by 20%, 15%, and 50%, respectively, by 2040 [21]. The electricity load of Ulaanbaatar city is calculated in three scenarios, with the lower scenario providing an annual growth value of 2.5%, the upper scenario 5.5%, and the average scenario 4.4%. In 2040, the load will be 1361 MW in the lower scenario, 2422 MW in the upper scenario, and 1950 MW in the average scenario. The draft Ulaanbaatar City Master Plan (UBMP) by 2040 is based on the average growth scenario. The total heat demand of buildings connected to the district heating supply system is estimated to be 3910 Gcal/h (4548 MW) by 2040, and the decentralized heat supply is estimated to be 1280 Gcal/h (1488 MW) [8].

3. Energy System Characteristics of the City of Ulaanbaatar

Ulaanbaatar is the capital of Mongolia. Ulaanbaatar is the coldest capital city in the world and has a distinct continental climate with short summers and long, cold, and dry winters [6]. The average annual temperature is −2.2 °C. Ulaanbaatar is the main socioeconomic, institutional, and political center of the country. It accounts for nearly 67% of the national gross domestic product (GDP). In 2020, Ulaanbaatar’s economic structure consisted of a 0.3% primary industry, 42% secondary industry, and 57% tertiary industry. The main sub-sectors are mining, services, finance, and enterprise. The population grew from 1.2 million in 2010 to 1.59 million in 2020 with an annual growth rate of 2.33% [22].

Ulaanbaatar’s energy sector is based on coal, which accounts for 93% of the total energy production [23]. The three main combined heat and power (CHP) plants of the Central Energy System are in Ulaanbaatar and export almost half of the electricity produced to other provinces. The main energy sources for Ulaanbaatar’s electricity and heat supply are CHP 2, CHP 3, CHP 4, the Amgalan power plant, small heat-only boilers (HOBs), and household stoves. Ulaanbaatar has the highest energy consumption per square meter compared to other cities in the world, which is due to the high heating demand during the eight heating months combined with highly subsidized energy costs [6].

The electricity demand of Ulaanbaatar in 2020 was 2859 GWh/year, which was 32% of the national electricity demand [8]. About 99% of all households in Ulaanbaatar are connected to the central power grid. Industry and businesses are the largest consumers, accounting for 60% of the total electricity demand, followed by ger areas (the ger area is a low-density residential area in the urban periphery that lacks basic urban services and has predominantly underserved lots with substandard housing) and apartments with 22% and 18%, respectively. In 2020, 49% of households lived in apartments and 51% in ger areas. Apartments and entities are connected to the central heating system. In Ulaanbaatar, there are more than 11,700 buildings that use thermal energy, and the total thermal load of these buildings was 2924 Gcal/h and the installed capacity was 2579 Gcal/h in 2020 [8]. Most households in ger areas are not connected to basic urban services and use coal stoves for heating and cooking. The city of Ulaanbaatar faces major challenges in the energy sector, resulting in it being one of the most polluted cities in the world [6]. Ulaanbaatar is also severely affected by climate change and is a hotspot for GHG emissions and pollution. Although Mongolia’s contribution to global GHG emissions is only 0.09%, its GHG emissions per capita are more than 2.7 times the global average [24].

The sustainable energy transition in the city of Ulaanbaatar is not discussed at both the national and local levels and is largely unexplored. In Ulaanbaatar, there is a large gap between the supply and demand of electricity and heat, making energy security the most sensitive issue. The Mongolian government’s policy and institutional framework is more focused on national and regional projects than on the city. Due to the national institutional arrangements, the city of Ulaanbaatar does not yet have a clear goal for a sustainable and low-carbon transition at the national and city levels. In addition, within the existing legal framework, the city of Ulaanbaatar plays a very limited role in the energy supply and industrial sectors. Because the energy system is dependent on coal, the role and concept of a sustainable energy transition is still in its early stages, and the opportunities, challenges, and pathways out are not yet known at the local level. However, the city can manage its other end-use sectors, such as buildings, public transportation, and other off-grid systems, within the current regulatory framework. As Ulaanbaatar is Mongolia’s largest energy producer and consumer at present, the city has great potential for renewable energy, especially solar and wind.

4. Methodology

Energy is linked to socio-economic activities and environmental issues. Energy management defines the proper use of available resources for sustainable development [13]. Long-term energy planning helps governments to achieve sustainable development goals. Accurate forecasting results allow the decision makers to make an optimal decision to meet future demands [12].

The MAED model was developed and introduced by the International Atomic Energy Agency to analyze alternative options for energy sector development and to forecast future energy demands. The MAED model is based on the improvements and modifications of the MEDEE model and adopted developing countries’ specifics in energy demand, which is the most appropriate model for developing countries. The model has a flexible framework in terms of the data, energy consumption patterns, sectoral disaggregation, and customer use. MAED can calculate energy demand at different levels, such as activity, subsector and sector, country, and city levels. For forecasting the future energy demand of the city of Ulaanbaatar, the MAED model was selected for the following rationales, which are listed in

Table 1. Rationales for the choice of the MAED model.

Specific Features of Developing Countries [14]	Ulaanbaatar Specific	MAED Model Scope [14]
Lack of data	Lack of disaggregated sectoral data at city level	Flexible and requires base year data
Uncertainties	Lack of long-term policy and strategy	Includes both conventional and renewable energy sources
Urban and rural divide	Apartments and dwellings in ger areas	Can define up to ten types of dwellings in urban and rural households
Transition from traditional to modern energies	High dependence on fossil fuels and untapped renewables	Can specify up to eight fuels; however, the energy used for energy conversion is not considered
Poor power sector performance	High energy loss of systems and outdated infrastructure	Covers demand sectors and bottom-up approach
Inadequate investment decisions and misdirected subsidies	Highly subsidized energy system and lack of investment	Model does not include price and elasticity information
Lack of appropriate models and institutions	Absence of local energy planning and modeling and lack of institutional capacity and human resources	The model runs on Microsoft Excel and can be easily installed on a PC running Windows, and low skills are required

.

Future energy demand is estimated using medium- to long-term scenarios of demographic, socioeconomic, and technological developments. Regarding the demand calculation and methodology, demand is calculated in terms of useful energy, and the final energy demand is determined based on the market penetration and end-use efficiency. In addition, demand is estimated by linking energy intensity to the level of economic activity. The demand of each subsector is determined and added to the total final demand by using a consistent accounting structure throughout. The main equations for the sectors of the MAED model can be found in Appendix A. Industry (agriculture, construction, mining, and manufacturing), transportation, service, and household are the main energy consuming sectors, and the total energy demand of each sector is divided into a large number of end-use categories in MAED. The scenario is divided into two sub-scenarios in MAED, namely, socio-economic and technology-based scenarios. The model consists of two modules: Module 1—MAED_D calculates the energy demand, which is the main input for Module 2, and Module 2—MAED_EL calculates the hourly electricity demand.

Module 1—MAED_D was used in the study for the energy demand projection to 2050 for the city of Ulaanbaatar. The structure of the MAED_D modeling procedure is described in Figure 1. Forecasting energy demand with MAED_UB involved the following steps:

Step 1:

Determine base and reference years. The base year of MAED_UB was set to 2020. The reference years of the model were chosen between 2020 and 2050.

Step 2:

Prepare and preprocess the data. Since MAED_D has about 250 input parameters, secondary data were gathered from various sources, including annual statistics; forecasts, such as population and economic growth from government policies; sectoral data; and raw data collected and processed from relevant government agencies and state and city institutions, surveys, and reports. Socioeconomic data for Ulaanbaatar for the base year are presented in Appendix A.

Step 3:

Development of the baseline scenario and other scenarios. The scenarios for sustainable energy transition in Ulaanbaatar city are shown in

Table 2. Assumptions of scenarios.

Socio-Economic and Technological Parameters	Scenarios
	BAU	REF	NDC	RM
Social and Economic Data
Population growth, annual average	2.7%	0.4% *	1.7% **	1.7% **
Urban population growth rate	Remains the level of the base year	5%	5%	10%
GDP growth, annual average	6% *	6% *	6% *	9.8%
Proportion of three industries (%)		Increase in secondary industry	Increase in secondary industry	Service sector more dominant
Primary industry	0.3%	0.3% *	0.3% *	0.5%
Secondary industry	39.7%	44% *	44% *	34.5%
Tertiary industry	60%	55.7% *	55.7% *	65%
Energy Intensities and Factors
Energy intensity - Motive power - Electricity specific - Thermal use	Remains the level of the base year	Declining at low rate of 10%	Declining at medium rate of 30%	Declining at high rate of 50%
Electric technology penetration	Remains the level of the base year	Up to 20%	Up to 30%	Up to 50%
Energy efficiency improvements	Remains the level of the base year	Low improvements up to 30%	Medium improvements up to 40%	High improvements up to 50%
Policy intervention	None	Municipal policy interventions	Municipal + national policy interventions	Extensive policy interventions
Enabling technologies	Remains the level of the base year	Advanced technology alternatives (please see Table 3)
Usage of renewable energy resources	Remains the level of the base year	Remains the level of the base year 30%	Recommended renewable energy resources by 40% in households (ger (ger is a traditional Mongolian nomadic portable dwelling composed of wood and felt), detached homes)	Renewable energy sources by 50% in service and households (ger, detached homes)
Transport sector	Remains the level of the base year	Increasing share of electric vehicles by 20% in urban transportation	Increasing share of electric vehicles by 30% in urban transportation	Increasing share of electric vehicles by 50% in urban and intercity transportation

Notes: * UBMP 2040, ** NSO.

.

Step 4:

Scenario analysis and implementation of MAED_D. Analyzed and verified model results of final energy demands by sector and fuel type with base year consumption.

Step 5:

Results and policy implications. We compared the results with previous studies and proposed policy recommendations and further studies.

Sectoral decomposition: MAED_UB breaks down the sectoral contribution and composition of energy consumption by the four main sectors of industry, transport, households, and services. The main sectors are further subdivided into subsectors, with industry subdivided into agriculture, construction, mining, and manufacturing; transportation subdivided into freight, urban, and intercity; and households subdivided into urban and rural. Both the industrial and service sectors include the energy intensity of motive power, electricity, and thermal use, as well as efficiency and market penetration. The transportation sector includes energy intensity, modal split, and load factors for freight, intercity, and intracity transportation. The residential sector includes factors for space heating, water heating, cooking, air conditioning, and household appliances in urban and rural residential buildings. The service sector mainly considers space heating and air conditioning. In the case study, several items were excluded from the calculation that did not apply to Ulaanbaatar city, namely, pig iron production and feedstock, international transportation, and household air conditioning. As for the urban and rural populations, residential areas were attributed to the urban population and ger areas surrounding the city were attributed to the rural population. International flights were not included in the calculation because the airport was not located in the territory of Ulaanbaatar city. Road freight data were only available for the country’s border cities. However, road freight transport was included in the calculation. The city of Ulaanbaatar is the main railroad center, and both the southern and northern railroad lines pass through Ulaanbaatar. Therefore, rail freight was included in the calculation.

Scenario development: Future energy demand from 2020 to 2050 was simulated using four scenarios, namely, a business-as-usual scenario (BAU), a reference (REF) scenario, a scenario derived from Mongolia’s nationally determined contribution (NDC) targets, and a reinforced mitigation (RM) scenario. The BAU scenario describes how the situation at present would develop without additional policy interventions and technological advances. The REF scenario is based on the perceptions of key stakeholders and the existing policy framework and targets for the energy sector and climate change mitigation. In contrast, the NDC scenario considers the implementation of policies and measures aimed at mitigating GHG emissions under Mongolia’s NDC, including Ulaanbaatar, in addition to the existing policy framework [25]. The RM scenario relies on more aggressive policies, advanced alternative technologies, and renewable energy sources proposed by donors and various studies to limit global warming temperatures to below 2 °C and continue efforts to limit it to 1.5 °C.

Economic transition: Various population and economic growth projections developed in four scenarios. Population growth for the scenarios was based on the historical average growth consistent with the UBMP 2040 and National Statistics Office (NSO) projections. The GDP growth rate for the BAU, REF, and NDC scenarios was based on the existing policy framework, i.e., UBMP 2040 and Mongolia’s Vision 2050, which was assumed to be 6%. The RM scenario was the most ambitious scenario in terms of GDP and was based on the highest economic growth of UBMP 2040 (9.8%). The differences between these four scenarios are shown in Table 2. The distribution of urban and rural populations and the growth of population mobility, which also affect energy demand, were considered consistent in the REF and NDC scenarios. The share of primary industry remains the same in the BAU, REF, and NDC scenarios and increases slightly in the RM scenario. The service sector dominates in all scenarios, with the RM scenario showing the greatest increase in the service sector, among others.

Policy intervention: To date, the energy system in Ulaanbaatar is administered by national authorities [6]. Under the existing legal framework, Ulaanbaatar is only responsible for the decentralized heat supply of the city. UBMP 2040 is the key document that defines future energy needs and supply. However, UBMP 2040 is being updated at present and has not yet been approved. There are two long-term energy planning initiatives, the draft Energy Master Plan for Ulaanbaatar to 2050 and the draft LEEAP for Ulaanbaatar to 2040, both initiated by The German Agency for International Cooperation (GIZ) [6,21]. However, these documents have not been approved by the City Council. The main policy documents at the national level are the New Revival Policy of Mongolia adopted in 2022 and the 2050 Vision Long-term Development Policy of Mongolia in 2020, which envisages expanding the capacity of existing power plants and developing new energy sources.

The main policy objectives for energy and climate change are summarized as follows [24,26,27]:

-

Increase the share of renewable energy in the total installed power generation capacity to 20% by 2023 and to 30% by 2030;

-

Reduce heat losses by 20% in 2020 and 40% by 2030;

-

GHG emission reduction of 23% compared to the business-as-usual scenario by 2030;

-

Reduction in GHG emissions to produce 1 GCal of energy—0.49 tCO2e by 2023;

-

Reduction in internal energy consumption of CHP plants by 11.2% by 2023 and 9.14% by 2030;

-

Reduce electricity transmission losses by 10.8% by 2023 and to 7.8% by 2030.

Technological alternatives: The application of advanced technologies, electricity and renewable energy penetration, and the improvement of energy efficiency are key measures for the transition to a low-carbon energy supply and important drivers for the reduction in energy demand. As the GDP per capita increases, household and service sector energy consumption levels increase. However, efficiency improvements will offset the increase in energy demand in the residential, service, and industrial sectors included in the calculation. Energy intensity will decrease due to energy efficiency improvement initiatives. The technologies and renewable energy alternatives identified in the studies are proposed as technology alternatives to the three scenarios presented in

Table 3. Potential energy supplies and advanced technology alternatives.

Sectors	Potential Energy Sources	Energy Efficiency Improvements	Enabling Technologies
Electricity	- Photovoltaics (PVs) (roof top, ground-mounted) - Wind turbines - Waste incineration - New highly efficient CHP - Heat pumps: Air-source heat pumps Solar-assisted ground-source heat pumps	Efficiency improvements and expansion of transmission and distribution network	- Batteries - Thermal storage - Direct electric heaters
Heating		Expansion and efficiency improvements in the district heating network
Transportation			Electric vehicles
Households		Therma-technical retrofitting 3–10% annually	- Electric heaters - Installing metering devices in households - Energy efficient appliances - Clean stove
Service		- Therma-technical retrofitting - Energy efficient buildings	- Energy efficient appliances - Installing metering devices in buildings
Industry		Expansion and efficiency improvements	Direct electric heating

, with the exception of BAU [6,9,21,24,28,29,30].

5. Results

5.1. Final Energy Demand

The energy demand forecast was conducted under four different socioeconomic growth scenarios, such as BAU, REF, NDC and RM, between 2020 and 2050 at MAED_D. The result of the final energy demand shows that the BAU scenario has the highest energy demand, which is about 550 PJ by 2050, i.e., the energy demand has quadrupled compared to the base year 2020, as shown in Figure 2. The REF scenario is the scenario with the second-highest energy demand and follows the NDC scenario. The RM scenario has the lowest energy demand compared to the other scenarios, amounting to about 264 PJ by 2050, a doubling from the base year. This is due to high electricity penetration and energy efficiency measures and energy intensity reductions in all sectors. The RM scenario assumes the highest GDP growth and urban population growth. However, this scenario dramatically increases the energy efficiency and low-carbon technology penetration, which reduces the final energy demand.

Final energy demand per GDP is shown in Figure 3. Final energy demand for the RM scenario is reduced by 2.7 times from 139.2 MJ/USD to 51.5 MJ/USD by 2050 compared to the base year.

5.2. Final Energy Demand by Sectors

In terms of the sectoral final energy demand, the BAU scenario is the scenario with the highest final energy demand in all sectors, except the household sector shown in Figure 4. The RM scenario has the lowest demand in all sectors among the other three scenarios. Final energy demand in the industry, transport, and services sectors increases 5.2 times by 2050 compared to 2020 in the BAU scenario. In the residential sector, it increases by 1.4 times. In the RM scenario, the final energy demand increases 2.6 times in the transport and services sectors and 2.2 times in the industrial sector. In the household sector, the demand reduces by 1.12 times. This is due to significant improvements in dwelling heat losses resulting from the high penetration of electricity and renewables in space heating and water heating, fossil fuel reductions and efficiency improvements, and improvements in clean cooking in ger areas.

5.3. Final Energy Demand by Energy Forms

Ulaanbaatar’s energy system includes four main types of energy commodities, presented in Figure 5 and Figure 6. Among the forms of energy, fossil fuels have the largest share, followed by motor fuels, district heating, electricity, and renewable energy. In the BAU scenario, the demand for motor fuels, fossil fuels, and district heating is the highest by 2050, and the reverse is true for electricity. However, in the RM scenario, it is the opposite compared to the BAU scenario. In the RM scenario, energy efficiency increases for all energy forms. Due to high electricity penetration and renewables, the amount of coal is reduced 7.2 times between the base year and 2050 in the RM scenario. Therefore, the final energy demand for electricity is the highest and increases 7.4 times in the RM scenario, while the demand for other fuels is the lowest.

The share of renewables in the power grid was about 5% in 2020. All three scenarios increased the share of renewables primarily in the residential and industrial sectors. The share of renewables increased by 12% in the REF scenario, 20% in the NDC scenario, and more than 25% in the RM scenario. The share of renewables in the power grid will gradually increase and reach 26 PJ by 2050 under the RM scenario. The potential renewable energy sources are estimated in the draft Energy Master Plan, which foresees a potential of 7141 MW from rooftop-mounted PVs, 1500 MW from ground-mounted PVs, and 600 MW from wind turbines in 2050 in Ulaanbaatar [6].

6. Discussion

Four scenarios were developed based on existing local and national socioeconomic and energy policies, as well as more ambitious policy and technology measures proposed and recommended by various donor initiatives in the MAED_D model. The final energy demand was calculated to be 548, 460, 334, and 264 PJ in 2050 for the BAU, REF, NDC, and RM scenarios, respectively, compared to 135 PJ in 2020. The results show that the high penetration of electricity and renewable energy, energy efficiency measures, and energy intensity reduction in all sectors can significantly reduce the future energy demands of Ulaanbaatar city.

Of the four sectors, the household sector had the highest energy consumption in the BAU scenario in 2020 at 32%, followed by industry at 30%, services at 25%, and transport at 13%. In the RM scenario, the share of the industrial sector increased to 34%, the service sector to 33%, the transport sector to 18%, and the share of the household sector decreased to 14% by 2050. Among the five main energy forms, fossil and motor fuels had shares of 37% and 35%, respectively, in the BAU scenario in 2020. Electricity and solar energy accounted for the smallest share of total energy forms, 7% and 1%, respectively. However, this ratio changed in the RM scenario, as the share of motor fuels increased to 48%, electricity to 26%, district heating to 13%, renewables to 10%, and fossil fuels to 3% by 2050, as shown in Figure 7.

The UBMP 2040 assumes a 2.75-times increase in electricity demand to 21 PJ by 2030 and 28 PJ by 2040 [8]. According to the draft Energy Master Plan, it is assumed that Ulaanbaatar’s total electricity demand will increase by 63% from 7 PJ in 2016 to 11.4 PJ by 2050. Total heat demand is assumed to increase by 36% from 36.5 PJ in 2016 to 50 PJ in 2050 [6]. Electricity demand in the BAU scenario was estimated in our study to increase from 9.3 PJ in 2020 to 18.3 by 2030, 25 PJ by 2040, and 35 PJ by 2050. The results of the BAU scenario are more consistent with the UBMP 2040. However, due to the high penetration of electricity in all sectors, the RM scenario showed a 7.4-times increase in electricity demand to 69 PJ by 2050. District heating demand in the BAU scenario was estimated to increase 3.4 times, i.e., 92.5 PJ by 2050 compared to 27.3 PJ in 2020. The RM scenario showed a 1.3-times increase in district heating demand, i.e., 35 PJ by 2050 due to the reduction in coal-based district heating.

Within the framework of the existing local and national policies and measures, the expansion of existing CHP and thermal power plants with an installed capacity of 1400 MW is planned. New liquefied petroleum gas (LPG) thermal power plants are planned in sub-centers in ger areas. All HOBs are planned to be phased out by 2025. Losses in power generation, transmission, and distribution are to be reduced to 8–9% by 2030. New smart and advanced technologies will be introduced to reduce emissions and improve energy efficiency. An important measure in the transport sector is the replacement of electric busses in public transport, which should reach 20% by 2030. In the building sector, improving energy efficiency is the government’s main policy, aiming to reduce the energy consumption levels of residential and public buildings by 15–50% by 2040 and reduce emissions from buildings by 60%. The industrial sector will introduce low-emission technologies. In addition to the existing policies, several studies have been conducted to assess potential renewable energy sources and capacity, determine readiness levels, and recommend low-carbon technologies for climate change mitigation at the national and local levels. Since Mongolia and Ulaanbaatar have great potential for solar and wind energy, rooftop- and ground-mounted solar PVs, wind turbines, waste incineration (53MW of electricity and 19 MW of heat), heat pumps (7744 MW), and new advanced CHP plants (573 MW) are recommended as energy sources and installed capacities by 2050. Batteries, thermal storage, and direct electric heating (3966 MW) are suggested as enabling technologies [6,9].

The study was a pilot study using the MAED model to forecast total and comprehensive energy demands, including the transport sector in the city and Mongolia. The study had several limitations. The most challenging were the data. First, most of the data used in MAED were not available at either the national or city levels. For example, energy intensity data for some sectors were not even available at the national level. While there are annual energy statistics published by the national authority, most of these data could not be used directly for MAED due to a lack of data measures and data required for MAED. Second, the data were not disaggregated to the city level. For example, the data on freight transportation were not available at the city level. Third, if the data were available, the sectors could be broken down into subsectors. For example, the household sector could be further subdivided into apartments and home types, such as single-family homes, precast panel apartments, apartments connected to basic infrastructure, and gers and single-family detached homes without a basic infrastructure. Therefore, future studies on different application areas of the MAED model at the city level could be considered: (i) data analysis for long-term energy planning models; (ii) improving the forecasting of future energy demands in subsectors; (iii) detailed analysis of potential renewable energy sources and advanced technologies and their application; and (iv) application of the MAED_EL module for electricity demand.

The MAED model had a flexible framework in terms of energy consumption patterns, sectoral breakdown, and customer usage; however, some limitations were identified in the MAED_UB model: (i) the MAED model lacked a detailed breakdown and specific parameters for renewable energy sources and technologies. The model considered all renewable energy sources as a single category, i.e., solar thermal. Different renewable energy sources, such as wind, solar, hydro, geothermal, etc., and technologies had different efficiencies, installation costs, lifetimes, and operating characteristics, as well as variations due to the weather conditions and other factors. (ii) The MAED model was recommended as an appropriate model for developing countries due to the lack of data availability or data collection systems. However, the MAED model was a data-intensive and detailed input data model, and the accuracy and reliability of the MAED model was highly dependent on the quality and availability of the input data. (iii) The MAED model focused on analyzing and forecasting energy demand and did not take into account supply side considerations. (iv) The MAED model may not explicitly consider the potential impact of policies, such as energy efficiency regulations, renewable energy incentives, and carbon pricing, as well as behavioral changes that can significantly affect energy demand.

7. Conclusions

Energy security and meeting growing energy demands are key challenges for the city of Ulaanbaatar. Therefore, decision makers at the national and city levels need to prioritize long-term energy planning and forecasting, as well as the transition to low-carbon energy. This study was the first to attempt to use the MAED model for long-term energy demand forecasting for the city of Ulaanbaatar in Mongolia. Several aspects need to be considered for further long-term energy planning and energy demand forecasting, as well as for future studies.

-

The transport sector was not considered as part of the energy sector in the existing regulatory framework. Therefore, the transport sector should be considered and integrated into the energy system, and its regulation under the Mongolian Energy Law is a crucial step toward a more comprehensive and sustainable energy policy.

-

Comprehensive and internationally recognized standardized data and collection guidelines for the energy sector need to be developed and collected by relevant institutions. Stakeholders involved in energy planning and policy should invest in the data collection, establish reliable data management systems, and collaborate with experts and relevant institutions to ensure the availability of accurate and up-to-date data. All the sectoral data must be disaggregated and collected by sector and subsector at the city level. In addition, continuous improvements in the data collection methods and advances in the data analysis can further increase the accuracy and relevance of the model.

-

Detailed and feasibility studies are needed to identify the potential renewable energy sources in Ulaanbaatar. Assessing the availability and potential of various renewable energy sources, including solar, wind, biomass, and geothermal, as well as detailed studies of solar radiation, wind patterns, biomass availability, and geothermal potential are critical to determining the suitability of each resource for Ulaanbaatar. In addition, it is critical to the stability and reliability of the energy system to study the technological feasibility of deploying renewable energy systems in the local context and to assess the existing energy infrastructure and grid capacity to define the feasibility of integrating renewable energy sources into the grid.

-

By prioritizing and investing in the low-carbon energy transition and implementing innovative and transformative low-carbon solutions, cities can play a central role in addressing climate change and promoting sustainable urban development. Ulaanbaatar can realize its renewable energy potential, transition to a more sustainable energy system, and contribute to Mongolia’s overall climate and energy goals.

-

Based on the results and the different scenarios of the final energy demand, both national and local authorities can analyze the energy consumption of each sector and introduce and prioritize low-carbon technologies and solutions and include them in policy documents with financial and political incentives. For example, the agricultural sector has the greatest energy intensity in motive power compared to the other sectors, which has a high potential for energy savings.

The city of Ulaanbaatar has great potential to reduce its energy consumption and dependence on coal and replace it with clean and renewable energy. To do so, it needs to define its optimal energy sustainability based on long-term energy demand planning, increase electrification and the use of renewable energy sources, implement energy efficiency measures, switch to clean fuels, and apply advanced and energy-efficient technologies in all end-use sectors. Therefore, the city needs to play a leadership role in its energy system to make the transition to sustainable energy.