Analysis of Cost-Optimal Renewable Energy Expansion for the Near-Term Jordanian Electricity System

Abstract

Jordan is affected by an ever changing environment in the midst of climate change, political challenges, a fast growing economy and socio-economic pressures. Among other countries in the Middle East and Northern Africa, Jordan is facing a number of electricity related challenges, such as a rising energy demand, high dependency on fossil fuel imports and management of local, fossil and renewable resources. The paper presents an analysis based on an open source optimisation modelling approach identifying a cost-optimal extension of the Jordanian electricity system with growing demand projections until 2030 utilising pumped hydro energy storage and determining the costs of different CO₂ mitigation pathways. The results highlight the large potential of renewable energy for the cost effective, environmentally friendly and energy independent development of the Jordanian electricity sector. A share of up to 50% renewable energy can be achieved with only a minor increase in levelised cost of electricity from 54.42 to 57.04 $/MWh. In particular, a combination of photovoltaic and pumped hydro storage proved to be a superior solution compared to the expansion of existing shale oil deployments due to high costs and CO₂ emissions. Aiming for a more than 50% renewable energy share within the electricity mix calls for substantial wind energy deployments. In a system with a renewable energy share of 90%, wind energy covers 45% of the demand.

1. Introduction and Background

Jordan is, in the midst of global warming and socio-political pressures, at an energy crossroads. Despite being in the middle of several oil-rich countries in the Middle East, Jordan is struggling to increase energy independence, being reliant almost entirely on fossil fuel imports. Despite having substantial renewable energy resources to increase energy independence and reduce greenhouse gas emissions, the most recently published energy strategy for 2040 [1] is more than conservative regarding their aims to increase the renewable energy share. Prior to the Arab Spring, Jordan relied almost entirely on natural gas imports from Egypt for electricity generation, which were disrupted in 2013 [2]. To satisfy energy demands, Jordan consequently switched to a petroleum based system. The government introduced substantial fuel subsidies to meet the increased costs and make energy available and affordable for the population [3], resulting in major governmental debt.

Between 1960 to 2011, six regional conflicts had direct or indirect effects on the energy sector in Jordan, namely, the Six Day War, the Lebanese Civil War, the Iraq–Iran War, the First Gulf war, the invasion of Iraq and the Egyptian revolution [4]. More recently, Jordan has been facing additional challenges concerning the energy and water sectors, such as a low level of foreign investment and substantial population growth due mainly to migration from war stricken Syria [5,6]. Unfortunately, there is a lack of literature dealing with the roles of foreign policy and politics and energy security in the case of Jordan. This rapidly changing and highly uncertain environment underlines the necessity for highly flexible energy system analysis tools to swiftly adjust to new circumstances.

1.1. Electricity Supply and Demand

Regardless of being in the midst of several oil-rich countries, Jordan struggles to secure its own energy resources for improved energy independence. Only recently, local oil shale resources were exploited, and a minor share of locally extracted natural gas was introduced to the system [7]. Table 1 introduces the installed capacities of Jordan’s electricity system as of 2018 and the near-term forecast for 2023. In 2011, 97% of Jordan’s energy needs were covered by oil and gas, consuming 19% of the Jordanian gross domestic product (GDP) [8] and only 2% were covered by renewable energy sources [9]. More recent, figures show that 19% of the installed capacity is covered by renewable energy power plants with 10.7% of the electricity generation of 2018 being covered by renewable energy sources [7]. Within the National Energy Masterplan for 2007–2020 [10] a reduction of energy dependency from 82% to 40% in 2020 was envisioned, which was not achieved, as Jordan is still importing 94% of its oil and gas to meet energy needs [11]. In 2018, 15% of the total electrical power consumption was used for water pumping, 45% in the residential sector, 22% in the industrial sector, 15% for commercial purposes and 2% for street lighting [12]. However, Jordan holds large renewable energy potential [5].

In 2018 the Jordanian peak load amounted to 3205 MW, which meant a decrease by about 3.4% from 3320 MW in 2017 [7]. NEPCO (National Electric Power Company) expects an increase of 1.9% by 2019 and a further 3% increase annually between 2019 and 2040 [7,12]. This means an electricity demand increase from 20,143 GWh to 38,261 GWh in 2040. Omary et al. [13] analyse the peak power demand development in three scenarios. The business as usual scenario assumes that the demand for electrical energy will grow continuously according to the growth of the last decade, reaching 25.3 TWh in 2030. The upper scenario assumes a higher increase with 30.3 TWh in 2030. The lower peak demand development path assumes a much lower demand of 15.8 TWh. Earlier studies examining potential future energy systems for Jordan, such as [14], expected an electricity demand of 106 TWh in 2050. This is due to the demand development between 2007 and 2013, showing a steady growth of electricity loads with an actual increase in consumption of 6.8% on average per year in the mentioned time period [8,12,15]. This was predicted to resume with a projected growth rate of 7.4% annually between 2014 and 2020 within the Master Strategy for Energy in Jordan [10], leading to an overestimation of the current demand. However, the most recent study from the Jordanian University of Science and Technology from 2019 [16] estimates a higher demand of 82.4 TWh for 2050 partly because of an electrification of other sectors. The unexpected demand decrease between 2018 and 2019, however, lead to a halt in the development of renewable energy projects. Before 2018, Jordan was progressing with the installation of renewable energies, becoming a leader in the Middle East on renewable developments. However, Jordan suspended renewable auctions and licenses for projects of 1 MW as of January 2019, due to concerns related to grid capacities [17]. Even considering a strictly fossil fuel based system, the future rising demand needs to be addressed, giving more stress to the grid, invalidating the argument of lacking grid capacity. Additionally, Jordan cancelled the tender for the first planned electrical storage project for renewable energy in 2020, inter alia due to the uncertain financial situation because of the global pandemic.

1.2. Strategies and Targets

A number of energy strategies were developed in Jordan: among others, the most relevant are the Energy Strategy 2030, the Energy Sector Strategy 2015–2025, the National Renewable Energy Action Plan, the National Energy Efficiency Action Plan and the Climate Change policy. However, some aims and visions are contradictory and incompatible, and previously set targets and goals were revised and neglected. For example the National Master Strategy of the Energy Sector for 2007–2020 and the National Strategy for the Development of Renewable Energy Resources stated the aim of 10% electricity generation based on renewable energies (wind and solar) by 2020, increasing to 20% by 2025. However, the latest energy strategy released in 2019 for 2018–2030 [1] aims in the baseline scenario at 21% for 2030, which indicates no evolution of the previously developed strategy. The penetration of renewable energy (RE) in the primary energy supply is predicted to increase from 3% in 2017 to 5% in 2020, reaching 6% in 2025 with no further increase up to 2030 in the reference case scenario [1]. The most ambitious scenario, increased sustainability, aims at 11% renewable penetration rate in 2030. Still, the energy import dependency will be as high as 73% in this scenario and between 92–94% in the reference case and business as usual scenario. As a least cost solution the share of renewable energy will not exceed 2.6 GW by 2030 (38% of installed capacity) with respect to the 2.4 GW of existing permits, and 5.7 GW by 2050 (47% of installed capacity). Contrary to the previous energy strategy and annual reports, which included the development of a nuclear power plant with a capacity of 220 or 660 MW being operational by 2026 [7], the revised strategy [1] does not foresee the development of nuclear energy.

Regarding CO₂ emission reduction targets, the revised energy strategy [1] envisions a CO₂ reduction of 10% by 2030; however, it fails to name a reference year. This goes along with the Intended Nationally Determined Contribution (INDC) [18], which aims at reducing greenhouse gas (GHG) emissions by 14% until 2030, again lacking a reference year. The mentioned 14% will be unconditionally full-filled by the country’s own means at a maximal 1.5% reduction compared to a business as usual scenario. In comparison, the European Union’s (EU) target is a reduction by 40% in 2030 compared to 1990 levels, and carbon neutrality by 2050 [19]. The temporary freeze in new renewable energy projects puts even this unambitious target at risk.

1.3. Research Question

The situation in Jordan puts strong emphases on energy independence and energy security because of the political and economical difficulties in the region. The research questions in this paper deal with the techno-economic assessment of the mid-term feature (2030). Therefore, we present an open source model based on the Open Energy Modelling Framework (oemof) [20] for the Jordanian electricity system. With the model the following research questions will be answered: (1) What is the cost-optimal mix, based on the current system, to meet the future electricity demand? (2) How can the future electricity demand be met by renewable energies in combination with pumped hydro and battery storage? (3) What are the costs of different RE shares in the electricity system?

2. State of the Art

2.1. Future Scenarios

The most recent study depicting the current Jordanian electricity system, as of 2018, underlines the number of challenges Jordan is facing, especially considering the current and coming energy demands [13]. The study does not offer any scenarios where energy storage is utilised, but does emphasise and suggest that the use of renewable energy resources could play a major role in a carbon relieved and more energy independent Jordanian energy system for 2030. Another recent study on electricity generation for Jordan was conducted by the University of Jordan, identifying mainly different scenarios in the face of two main issues, which are economic reasoning and geopolitical uncertainties [21]. Using GAMS, the study identified Combined Cycle Gas Turbine (CCGT) units mainly fired by natural gas in combination with PV and wind as the optimal choices with 70%, 19% and 11% shares respectively in 2018, changing to 10%, 71% and 19% in 2035. Dawoud et al. [21] recommend providing integrated storage options, without introducing concrete possibilities. Researchers from the school of energy systems with LUT Unjversity, Finland, [22] conducted an in-depth analysis of energy security of a 100% renewable energy transition in Jordan by 2050, projecting renewable electricity generation to increase from 0.1 TWh in 2015 to 110.7 TWh in 2050, 92% being covered by solar energy. Therefore, levelised cost of electricity (LCOE) develop from 78 EUR/MWh in 2015 to 61 EUR/MWh in 2050. For the calculation of scenarios an expansion model with 5 year time periods and an hourly resolution within each year is applied. The study recognises the importance of energy storage within a renewable system, introducing battery storage from 2025 onwards, with an installed capacity of 1 GWh increasing to approximately 67 GWh in 2050. Additionally, a compressed air energy storage (CAES) is included in the system in 2030, with a capacity of 31 TWh in 2050. Another study by the Jordan University of Science and Technology from 2019 [16] established various scenarios using EnergyPlan and LEAP, with one being 100% renewable, while others integrated natural gas, oil shale and nuclear power. The 100% renewable scenario was introduced with a high share of concentrated solar power (CSP), 10.6 GW, wind power of 4.5 GW and 25 GW of PV to cover the predicted demand of 2050 (82.4 TWh respective 14,350 MW peak load), and introduced a 90 GWh storage system to meet dispatchability problems. Kiwan and Al-Garibeh [16] found that the 100% is also economically feasible, with cumulative expansion cost of the renewable system amounting to $60 Billion compared to $52 Billion for the conventional system.

Within the MENA-Select (Sustainable Electricity Trajectories) project, participatory scenarios for the future Jordanian energy system for 2050 were established with local stakeholders [14]. In the other participating countries, Tunesia and Morocco, 100% renewable energy scenarios were considered and investigated; however, Jordanian stakeholders did not explore this possibility. The lowest CO₂ emissions were achieved within the no imports scenarios, reliant heavily on wind and PV (15 GW, 25 GW) as well as CSP (20 GW) and oil and gas (5 GW and 4 GW). Here, the largest energy storage (batteries) was modelled, with a capacity of 18 GW and an energy capacity of 40 GWh. Although this scenario is by far the most expensive, it was ranked the most preferable by local stakeholders due to the increased energy independence. However, due to changing developments, these scenario are not suitable to give guidance for the near and mid-term future. In another consecutive study, IIASA (International Institute for Applied Systems Analysis), among others [23], identified that energy security is preferable for all stakeholders over environmental concerns, which might be why a 100% renewable option is not as relevant for Jordan as for other countries.

2.2. Pumped Hydro Storage

The possibility of a pumped hydro storage system for Jordan was analysed within the Renewable Energy and Energy Efficiency Program for Jordan [24,25]. Pumped hydro storage (PHS) can facilitate a smoother integration of renewable, volatile energy sources into the national electricity system, if geographical features are beneficial. In Jordan, out of ten water reservoirs, three were identified to hold potential for pumped storage plants, namely, Mujib, King Talal and Wadi Arab. Mentioned here is the need for further studies to investigate the energy storage demand within the energy system, to verify assumptions. The study, however, did not analyse the integration of a pumped hydro storage into the Jordanian electricity system. The current energy strategy [1] advises in the increased sustainability, minimum dependency and rational use of energy scenario, a PHS of 220 MW to be introduced by 2025, to avoid renewable energy curtailment. Furthermore, the strategy selects Mujib as the only cost effective option. Generally speaking, a number of studies have been conducted that have identified the benefits of hybrid pumped hydro and battery storage for renewable energy based power systems, e.g., most recently [26]. In the Jordanian context, a number of studies have analysed in detail the renewable energy potential, such as a study by the Tafila Technical University [27] revising the renewable situation in Jordan in 2005. Here, a strong case for pro renewable energy was made, regarding energy security, energy independence and potentially lowered costs due to less operation and maintenance, as well as the environmental benefits in contrast to conventional energy sources. A more recent study of 2019 [28] explored the possibility of a combination of wind and pumped hydro storage within the Jordanian energy system. The team from the Yarmouk University, the University of Jordan and Texas A&M University used a Matlab optimisation toolbox to find the cost-optimal solution, showing that a combined wind and hydro storage system is economically, environmentally and technically more efficient than conventional power generation with CO₂ emissions and conventional grid energy purchases being reduced by almost 25%.

2.3. Contribution

As the State of the Art section shows, several modelling and scenario efforts have been made around the future Jordanian energy system. However, there is not an open source energy system modelling approach for Jordan, nor have the defined research questions been addressed. To our knowledge, no studies analysed how the existing energy system can be extended by renewable energy sources in combination with pumped hydro and battery storage to meet the expected rising energy demand in Jordan. In fast changing environments, open source models with open data can be of great value to adapt in a short manner. Additionally, the presented open source model can be used to assess similar research questions for any other country. Therefore, the presented work not only contributes to the scientific debate on decarbonisation of energy systems for climate change mitigation in Jordan, but also builds an important bridge for capacity building and development cooperation for other countries. The tool can facilitate a discussion among different sectors, e.g., the water and energy sectors, to identify joint solutions for common problems.

3. Mathematical Model

The developed and applied model is a linear (mixed-integer) optimisation model for the Jordanian electricity system. It is based on the open source package oemof-tabular [29]. In the following, endogenous (optimisation) variables are shown in bold to differentiate between these and exogenous model variables. The model minimises total operational cost for the time horizon T and all units $u\in U$, and annualised investment cost of all units $i\in I$, along with storage investments of all storage $s\in S$ for the Jordanian electricity system. Elements of the sets for the scenarios are listed in the Appendix A. The respective objective function is given below in Equation (1). The implemented model as well as the input data are provided in the Supplementary Material.

(1)$\begin{array}{c}\hfill \min :\stackrel{\mathrm{operational}\cos \mathrm{t}}{\overbrace{\sum _{t\in T}\sum _{u\in U}{c}_u^{opex}{\mathbf{p}}_{u,t}}}+\stackrel{\mathrm{power}\mathrm{inv}.\cos \mathrm{t}}{\overbrace{\sum _{i\in I}{c}_i^{capex,p}{\mathbf{p}}_i^{nom}}}+\stackrel{\mathrm{energy}\mathrm{inv}.\cos \mathrm{t}}{\overbrace{\sum _{s\in S}{c}_s^{capex,e}{\mathbf{e}}_s^{nom}}}\end{array}$

The operational costs are calculated based on the efficiency ${\eta }_u$ of a unit u and its fuel cost $c_u^{fuel}$ according to Equation (2). Annualised investment costs $c^{capex}$ are calculated based on the lifetime n, weighted cost of capital (WACC) i and specific investment cost of a technology $CAPEX$, along with the fixed operation and maintenance cost $FOM$ in Equation (3). The scenario specific values for this study are found in Table 2 in the next section.

(2)$\begin{array}{cc}\hfill c^{ope{x}_u}& =\frac{c_u^{fuel}}{{\eta }_u}\hfill \end{array}$

(3)$\begin{array}{cc}\hfill c^{capex}& =CAPEX·\frac{(\mathrm{i}·{(1+\mathrm{i})}^n)}{({(1+\mathrm{i})}^n-1)}·(1+FOM)\hfill \end{array}$

Demand must equal the sum of supply of all producing units, as described in Equation (4). Note that in the case of the storage units, p can also take negative values when the storage is charging.

(4)$\begin{array}{c}\hfill \sum _{u\in U}{\mathbf{p}}_{u,t}=d_t+{\mathbf{p}}_t^{excess}\forall t\in T\end{array}$

For all investment units, the supply is limited by the installed nominal power p${}_i^{nom}$ described in Equation (5).

(5)$\begin{array}{cc}\hfill 0\le {\mathbf{p}}_{i,t}\le {\mathbf{p}}_i^{nom}& \forall i\in I,t\in T\hfill \end{array}$

(6)$\begin{array}{cc}\hfill {\underline{p}}_i\le {\mathbf{p}}_i^{nom}\le {\overline{p}}_i& \forall i\in I\hfill \end{array}$

The energy storage balance in Equation (7) is applied for all modelled storage types. The balance includes standing losses ${\eta }^{loss}$ as well as charge and discharge efficiencies ${\eta }^{in/out}$.

(7)$\begin{array}{c}\hfill {\mathbf{e}}_{s,t}={\mathbf{e}}_{s,t-1}·{\eta }_s^{loss}-\frac{{\mathbf{p}}_{s,t}^{out}}{{\eta }_s^{out}}+{\mathbf{p}}_{s,t}^{in}·{\eta }_s^{in}\forall s\in S,t\in T\end{array}$

Additionally, the power of the storage is limited by the optimised nominal power shown in Equation (8).

(8)$\begin{array}{cc}\hfill -{\mathbf{p}}_s^{nom}\le {\mathbf{p}}_{s,t}\le {\mathbf{p}}_s^{nom}& \forall s\in S,t\in T\hfill \end{array}$

For all RE technologies, i.e., PV and wind, the power output is determined by Equation (9) where $c_t^{profile}$ is the time-dependent normalised generation profile of the unit $i\in I$. The profile data can be obtained from measurements, calculated from re-analysis weather data or directly obtained from databases such as renewables.ninja [30,31].

(9)$\begin{array}{cc}\hfill {\mathbf{p}}_{i,t}=c_{i,t}^{profile}{\mathbf{p}}_i^{nom}& \forall i\in I,t\in T\hfill \end{array}$

Analogously to Equations (5) and (6), the energy storage level and its maximum investment level are bounded as shown in Equations (10) and (11).

(10)$\begin{array}{cc}\hfill e_s^{min}·{\mathbf{e}}_s^{nom}\le {\mathbf{e}}_{s,t}\le {\mathbf{e}}_s^{nom}& \forall s\in S,t\in T\hfill \end{array}$

(11)$\begin{array}{cc}\hfill 0\le {\mathbf{e}}_s^{nom}\le {\overline{e}}_s& \forall s\in S\hfill \end{array}$

For all conventional units $c\in C$, upper and lower limits for the total energy supply over the time horizon T can be bounded with Equations (12) and (13).

(12)$\begin{array}{c}\hfill \sum _{t\in T}{\mathbf{p}}_{c,t}\ge {\underline{E}}_c\forall c\in C\end{array}$

(13)$\begin{array}{c}\hfill \sum _{t\in T}{\mathbf{p}}_{c,t}\le {\overline{E}}_c\forall c\in C\end{array}$

To model RE penetration within the system by an exogenously defined RE share an additional constraint is introduced. The renewable energy share is defined by Equation (14) by the share of conventional technologies $c\in C$.

(14)$\begin{array}{c}\hfill \sum _{t\in T}\sum _{c\in C}{x}_c^{flow}\left(t\right)\le (1-R{E}^{share})·c_l^{amount}\end{array}$

4. Scenario Assumptions

Within this study four different scenarios are modelled to analyse the future Jordanian electricity system. Based on NEPCO forecast, the demand for all scenarios is 28 TWh [7].

The BASE scenario, considering the existing power park of 2023 shown in Table 1, is a lower bound to the capacity expansion. The CONT scenario includes fossil fuel contracts for minimum gas consumption as well as operational constraints for the existing shale-oil power plant. The operation of the shale-oil unit is exogenous, set to 7500 h full load hours. For natural gas, 24 TWh_th annual gas consumption is set in the model. All other assumptions are the same as in the BASE scenario. As the energy independence in Jordan plays an important role, an AUT scenario, wherein only local resources can be utilised, has been added. Finally, the GRE scenario is an unconstrained electricity mix optimisation (greenfield planning approach). Therefore lower bounds on the investment of units were set; all costs and technical parameters were the same as in the BASE scenario. For all scenario setups, different shares of RE are modelled with Equation (14).

Costs and Technology Parameter

Table 2 summarises the cost and technology assumptions for all scenarios. For battery storage units a power to energy ratio of 1/6 was used; for PHS a ratio of 1/10 has been used in all scenarios. The PHS potential in this paper was derived from the work of [17]. According to the study, three (Mujib, Wadi Arab and King Talal) out of ten dams operated by the Jordan Valley authority are suitable for PHS installations. For these dams only an upper reservoir needs to constructed. Due to geological limitations, the aggregated PHS potential is restricted to 3750 MWh. Cost estimations for these PHS storage units are based on reference [34].

5. Results

5.1. Cost-Optimal Mix

The results of the scenarios for the cost-optimal mix are presented in Figure 1 and Figure 2. The CONT and BASE scenarios result in a similar technology mix with RE shares of around 33%. Compared to the current power park, almost no additional investment in conventional units is required to meet the future demand. Instead, PV is expanded by capacity of 4.27 GW while wind capacity is not expanded for the cost-optimal mix. The GRE scenario shows that without current restrictions, the optimal mix consists of 3.74 GW of CCGT followed by 4.47 GW of PV and 0.99 GW of GT. The only scenario where storage units are installed is the AUT scenario. With 375 MW the PHS potential is fully exploited, and an additional 1.28 GW of battery storage is installed. The RE share of above 60% is significantly higher compared to the other scenarios. In addition to substantial PV capacity of 6.78 GW, wind capacity of 3.58 GW and oil shale capacity of 2.88 GW of are installed.

Except for the AUT scenario with 10.27 TWh_el of shale oil based supply, most electricity is still supplied by conventional units in the cost-optimal mix. For the cost-optimal case with no constraints on the RE share, 16.41 TWh is supplied by CCGT in the BASE scenario. With the contracts applied, the oil shale unit supplies 3.53 TWh, which causes a drop in the CCGT supply to 14.01 TWh. In both cases, around 8.16 TWh is produced by PV units. Notably, emissions of the cost-optimal AUT scenario (9.42 million t), with a RE share of above 60%, are similar to the CONT scenario (9.62 million t) with a RE share of about 30%, as emission factors of shale oil are higher and efficiency is lower compared to CCGT units. Emissions within the BASE and GRE scenarios are lower with 7.93 and 8.27 million t respectively.

5.2. Varying Renewable Energy Shares

Figure 3 shows the installed capacities for 2030 for all scenarios with different RE shares. Detailed data are provided in the Appendix A. As described above, the cost-optimal mix in all scenarios already features a RE share 30% or above. Due to the lower bound on the gas consumption in the CONT scenario, higher shares of RE are not feasible within this setup. Compared to the status quo (2023), results show a significant increase in PV followed by wind investment in the BASE scenario to meet the increased demand of 28 TWh. In addition, minor investment in CCGT was chosen in the BASE scenario up to a RE share of 50%. This shows that due to the differences in marginal cost, additional CCGT investment is preferred instead of dispatching the shale-oil unit. PHS storage investment becomes relevant for RE shares of above 40% and the potential is fully exploited at shares above 50%. For up to 70% RE share, no additional storage than PHS is required to integrate the RE. Above 80% RE, investment in battery storage starts to increase significantly with over 3.15 GW installed capacity in the BASE-90 scenario and 3.82 GW in the AUT scenario.

Compared to the BASE scenario, a similar pattern with regard to installed capacities under different RE shares can be observed within the GRE scenario. However, in particular for shares above 80% RE, total conventional capacities are lower. Despite higher investment in shale oil, PV plays a bigger role than wind within the AUT scenario. In BASE-90 9.95 GW PV and 8.99 GW wind are installed compared to 10.89 GW PV and 6.93 GW in the AUT-90 case.

The energy supply, energy demand and corresponding CO₂ emissions are shown in Figure 4. For higher shares of RE, wind energy becomes more relevant and the need for additional battery storages increases significantly. In addition, limited (long) term storage options and missing transmissions to neighbouring countries cause high curtailment. In the BASE-90 scenario, over 35% of the RE production is curtailed. Due to higher storage capacities, curtailment is lower in the AUT scenarios.

With these results, three different stages within the system can be identified for the BASE scenarios: (1) low shares for up to 50% RE where PV supply is dominating; (2) medium share of RE between 60% and 80% where wind is higher than PV; and (3) high shares of above 80% where PV is equal to or more prevalent than wind. This shows the energy system’s dynamic. PV has lower single technology cost of electricity and integrates well until a certain level of RE penetration is reached. After this point, the system value of wind starts to increase because it can supply electricity when PV is not available. Despite excellent solar resources and low cost, up to about 50% of the electricity supply comes from wind for scenarios of 90% RE share. A similar pattern can be identified within the GRE scenario. In contrast, the AUT scenarios feature higher storage capacities and therefore also in all cases higher PV supply than wind.

A major difference between the CONT and the BASE scenarios is the resulting level of CO₂ emissions. Due to the shale-oil unit, emissions are significantly higher for the CONT scenario.

5.3. System Operation

Figure 5 shows the dispatch of units for the BASE-40 and BASE-80 scenario. Within the system displayed in Figure 5a, mainly PV supply is consumed during the day while in the evening peaks and during the night CCGT and GT units are providing electricity. Storage operation is not required to integrate the RE. In contrast, Figure 5b shows the electricity system with a 80% RE supply. Here, consequences of increased RE supply can be observed. Storage operation increases notably, integrating wind and solar supply during the day and shifting this electricity to the evening peak. In addition, the high excess of RE during the day is also clearly visible.

For the same scenarios, the aggregated state of charge of the PHS units is shown in the heat map plot in Figure 6. The PV integrating pattern with fully charged storage units during the day and empty storage units in the morning is visible. It can be observed that the storage is operated more intensively in the case with higher share of RE. During the summer months, the storage is fully charged during the whole day in the BASE-40 scenario, whereas this can not be observed in the BASE-80 scenario.

5.4. Costs

Figure 7 shows the levelised cost of electricity (LCOE) for varying RE shares. LCOE has been calculated by dividing the total annualised investments and operational costs by the electricity demand covered. Note that for renewable energy systems, additional costs occur for integrating the intermittent electricity into the system, as discussed in [35].

Clearly, the AUT scenario comes with the highest cost, as autarchy has a high price. However, as it already features a RE share of above 60% in the cost-optimal mix, the increase in cost towards a 90% RE share setup is rather low in relative terms. The second highest cost for the cost-optimal case can be found within the CONT scenario with oil shale supply. The green field planning scenario GRE highlights that a combination of CCGT and RE is more cost efficient. The LCOE of the BASE and GRE setup do not differ significantly. With regard to rising RE shares, it can be observed that up to 50% RE can be achieved with a very small increase in LCOE from 54.52 to 57.04 $/MWh in the BASE scenario. For a RE share of up to 90%, LCOE increases almost to values twice as high as that in the cost-optimal case. The effects of higher storage requirements and thus additional investment costs and high excess electricity with curtailment can be reasons for these figures.

While PV has the lowest LCOE as a single piece of technology, the majority of the investment costs in scenarios with high shares of RE are caused by wind energy deployments and battery storage. The distribution of investment costs depicted in Figure A1 in the Appendix A shows that a combination of technologies within a system that strives for high RE shares has a different value compared to a single technology solution.

6. Discussion

The disruption of natural gas supply in 2011 caused by the Arab spring proved the unreliability and instability of the Jordanian energy system. Ever since, the country has failed to increase energy independence. With abundant renewable energy resources, a combination of PHS and RE energy and efficient CCGT units is the most cost effective way for gaining increased energy independence and simultaneously reducing GHG emissions.

6.1. Comparison with Other Studies

The presented results indicate a high share of RE within a cost-optimal energy system to meet the increased energy demand in Jordan by 2030 compared to what is envisioned within existing strategies, such as in references [1,7,18]. The herein determined cost-optimal energy mix includes a share of above 30% of RE by 2030, in all scenarios. In addition, the presented study also identified a greater role of wind energy and PHS for the Jordanian electricity system, contrary to [13], which underlined the importance of RE, while neglecting PHS utilisation though. Compared to [21], analysing the cost-optimal energy mix combining CCGT with PV and wind (10%, 71% and 19% in 2035), the calculated results indicate a lower share of PV—29.2% within cost-optimal mix of BASE scenario.

The integration of PHS is vital to a system with high renewable energy shares. The necessity of PHS and long-term battery storage to increase the share of RE and increase energy independence is recognised by [21,22]. The authors of [22] aim for 100% renewable energy supply by 2050, integrating 1 GWh of battery storage in 2025 up to 67 GWh in 2050; in this study 0.69 GWh PHS is necessary within the BASE scenario, with a RE share of 40%, and only with a share of above 70% RE does the battery storage become necessary. A RE share of 90% within the BASE scenario makes 3750 GWh PHS storage necessary. Along with [1], PHS is identified as an option to avoid or limit curtailment of renewable energies.

Supporting [27], the results clearly show reduced CO₂ emissions within the cost-optimal setting, which includes in all scenarios 30% or more RE and the high potential of further reductions due to the high potential of renewable energies within Jordan. Additionally, [28] proposed the combination of wind and PHS to be economically, environmentally and technically more efficient than conventional power generation in regard to CO₂ emissions. This is supported in this study.

As analysed within [4], a diversification of energy generation can have beneficial effects on the energy security of Jordan. According to this analysis, most relevant measures are continuing the decrease of imported energy through the utilisation of domestic energy resources such as oil shale and renewable sources (wind and PV). While this is certainly true for wind and PV, shale oil is environmentally and economically not recommendable, as shown within the here presented analysis. Instead, under the assumption of a growing electricity demand, PHS in combination with PV and wind energy can provide a secure, environmentally beneficial and cost effective energy supply.

Results show that shares of up to 50% RE share can be achieved by a slight increase in LCOE. Due to required storage investment and curtailment of RE, there is a high increase of LCOE for shares of up to 90%. However, it is important to note that an integrated electricity system of countries in the MENA region could reduce system costs significantly, as shown by [36]. Such integration will also help to reduce curtailment. Similarly, smart sector integration of the water and electricity sector is another option to increase RE penetration in the Jordanian energy system. Jordan, as one of the water-scarcest countries on the planet [37], has a high energy demand for the water sector, which is likely to increase in the coming years due to increased need for water pumping because of lowered water levels as well as the need for desalination of water as an additional source of fresh water.

6.2. Limitations of the Study

The study applied an open source investment model to analyse the future Jordanian electricity system. However, results need to be read in light of the modelling limitations. First of all, it is important to note that no transmission to neighbouring countries and Jordan’s grid has been modelled. While the former can help to provide a solution with lower cost, due to reduced excess and lesser storage requirements, as discussed above, the latter can actually counteract these effects. In particular, curtailment and storage dispatch can be higher to keep the system balanced within the country on the distribution and transmission grid levels. Hence, storage units may be cost efficient within scenarios of shares below 40% RE.

Another important point is the cost-optimal dispatch based on perfect competition, where the existing contracts with independent power producers (IPP) need to be considered. While gas contracts have been integrated, additional contracts may exist that do not allow for a reduction of conventional power plant operation, and therefore limit RE expansion.

6.3. The Value of Open Source Tools

The context-specific boundaries, such as existing contracts, power plant characteristics and grid constraints, are important factors when modelling an electricity system. However, that information is not always available for scientists. In addition, political and economic dynamics can change fundamental assumptions, such as price and demand developments, in a short period of time. Therefore, open source approaches are of high value for further investigations. In addition to changes of basic assumptions and input data of this study, the model can be improved or extended. Among others, the applied Open Energy Modelling Framework (oemof) [20,29] provides the opportunity for detailed power plant modelling with minimum up and down times, part load efficiencies and linear optimal power flow grid modelling. Such functionalities could be integrated inside the developed model as well. The same holds for the PV and wind profiles, as [16] states, the exact renewable profiles which are technically feasible in Jordan have not been quantified yet.

7. Conclusions

The paper presents an analysis based on an open source optimisation modelling approach of the Jordanian electricity system in 2030. Results highlight and confirm the great potential of renewable energy for cost effective, environmentally friendly and more energy independent development in Jordan. Up to 50% renewable energy within the electricity system can be achieved with only a slight increase of levelised cost of electricity from 54.52 to 57.04 $/MWh. In particular, photovoltaic installations in combination with pumped hydro storage, as a low cost storage technology, seem to be a superior solution compared to the expansion of shale oil deployments due to high costs and CO₂ emissions. For higher shares of renewable energy, wind energy can play an important role, making up above 45% of the renewable energy supply in a 90% renewable energy based system.

However, high shares of renewable energy within the electricity mix require the analysis of long term storage options and grid expansion to neighbouring countries to avoid high costs as well as extensive curtailment of renewable energy. In addition, the water–energy sector cooperation using flexible desalination can be an important step to integrate renewable produced electricity and attenuate water stress at the same time. Within the transformation process, fossil fuel contracts pose a challenge, as they may hamper renewable energy expansion and increase integration cost. Modelling the scenario with existing long term gas contracts shows that renewable energy shares above 33% cannot be achieved, even under a growing electricity demand, by 2030. Therefore, strategic planning with a long term perspective is important for the Jordanian electricity system.

Supplementary Materials

The following are available online at https://github.com/znes/oemof-jordan/releases/tag/paper.

Author Contributions

Conceptualization, S.H.; methodology, S.H.; software, S.H.; validation, S.H. and F.D. and A.A.-S.; formal analysis, S.H.; investigation, S.H. and F.D.; resources, S.H. and A.A.-S.; data curation, S.H. and F.D. and A.A.-S.; writing—original draft preparation, S.H. and F.D. and A.A.-S.; writing—review and editing, S.H. and F.D.; visualization, S.H.; supervision, S.H.; project administration, S.H.; funding acquisition: S.H. and F.D.; All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ), grant number 81243459.

Acknowledgments

We acknowledge financial support for the article processing charge by the state of Schleswig-Holstein, Germany within the funding programme OpenAccess-Publikationsfonds. This work has been carried out during sabbatical leave granted to the co-author Ahmed Al-Salaymeh from the University of Jordan during the academic year 2019/2020. In addition, authors would like to thanks Clemens Wingenbach who was head of the project administration.

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:

CAES	Compressed Air Energy Storage
CC	Combined Cycle
DE	Diesel Engine
FOM	Fixed operation and maintenance
GAMS	General Algebraic Modelling System
GDP	Gross Domestic Product
GHG	Green house gas
IIASA	International Institute of Applied Systems Analysis
INDC	Intended Nationally Determined Contribution
LEAP	Low Emission Analysis Plattform
LCOE	Levelised Cost of Electricity
MENA	Middle East and Northern Africa
NEPCO	National Electric Power Company
GT	Gas Turbine
PV	Photovoltaic
oemof	Open Energy System Modelling Framework
RE	Renewable Energy
ST	Steam Turbine
WACC	Weighted average cost of capital

Appendix A. Results

Figure A1

Annualised investment cost within all scenarios in Billion US $.

[Figure omitted. See PDF]

Table A1

LCOE in US $/MWh.

	REF	40%	50%	60%	70%	80%	90%
CONT	58.76	-	-	-	-	-	-
BASE	54.52	55.15	57.04	61.04	69.92	83.68	101.83
GRE	48.37	49.26	50.91	54.79	63.31	75.06	91.87
AUT	95.59	95.59	95.59	95.59	96.28	100.91	109.58

Table A2

Annualised investment cost in million US $.

	Gas-cc	Gas-de	Gas-st	Gas-gt	Shaleoil-st	Wind-Onshore	Solar-pv	phs	Battery
CONT	138.27	70.9	55.67	7.02	118.09	64.77	261.93	0.00	0.00
BASE	144	70.9	55.67	3.07	118.09	64.77	262.27	0.00	0.00
BASE-40	140.08	70.9	55.67	3.07	118.09	135.65	288.44	5.56	0.00
BASE-50	138.27	70.9	55.67	3.07	118.09	303.45	293.54	10.43	0.00
BASE-60	138.27	70.9	55.67	3.07	118.09	479.53	332.55	30.16	0.00
BASE-70	138.27	70.9	55.67	3.07	118.09	795.27	365.10	30.16	21.22
BASE-80	138.27	70.9	55.67	3.07	118.09	881.53	460.03	30.16	346.41
BASE-90	138.27	70.9	55.67	3.07	118.09	878.64	610.81	30.16	826.25
GRE	201.19	0	0	36.84	0.00	0.00	274.11	0.00	0.00
GRE-40	184.21	0	0	37.23	0.00	113.16	304.40	19.88	0.00
GRE-50	167.89	0	0	43.4	0.00	277.35	304.91	30.16	0.00
GRE-60	154.26	0	0	52.39	0.00	480.34	333.36	30.16	0.00
GRE-70	136.23	0	0	53.36	0.00	738.83	372.42	30.16	77.48
GRE-80	102.57	0	0	29.37	0.00	807.10	477.83	30.16	410.32
GRE-90	64.84	0	0	31.69	0.00	841.63	622.36	30.16	855.80
AUT	-	-	-	-	724.27	349.56	416.02	30.16	335.34
AUT-70	-	-	-	-	686.15	460.42	451.38	30.16	406.25
AUT-80	-	-	-	-	626.15	673.96	511.25	30.16	543.04
AUT-90	-	-	-	-	471.19	676.66	668.73	30.16	1001.05

Table A3

Installed capacities in MW.

	Gas-cc	Gas-de	Gas-gt	Gas-st	Shaleoil-st	Solar-pv	Wind-Onshore	Hydro-ror	phs	Battery
CONT	2567	810	189	636	470	4267	663	12	0	0
BASE	2673	810	83	636	470	4272	663	12	0	0
BASE-40	2600	810	83	636	470	4698	1388	12	69	0
BASE-50	2567	810	83	636	470	4781	3106	12	129	0
BASE-60	2567	810	83	636	470	5417	4908	12	375	0
BASE-70	2567	810	83	636	470	5947	8140	12	375	80
BASE-80	2567	810	83	636	470	7494	9023	12	375	1320
BASE-90	2567	810	83	636	470	9950	8993	12	375	3150
GRE	3735	0	994	0	0	4465	0	0	0	0
GRE-40	3420	0	1005	0	0	4958	1158	0	247	0
GRE-50	3117	0	1172	0	0	4967	2839	0	375	0
GRE-60	2863	0	1414	0	0	5430	4916	0	375	0
GRE-70	2529	0	1440	0	0	6066	7562	0	375	295
GRE-80	1904	0	793	0	0	7784	8261	0	375	1564
GRE-90	1203	0	855	0	0	10,138	8615	0	375	3263
AUT	0	0	0	0	2882	6777	3578	0	375	1278
AUT-70	0	0	0	0	2730	7353	4712	0	375	1548
AUT-80	0	0	0	0	2492	8328	6898	0	375	2070
AUT-90	0	0	0	0	1875	10,893	6926	0	375	3816

Table A4

Energy supply and demand in TWh.

	Gas-cc	Gas-de	Gas-gt	Gas-st	Shaleoil-st	Hydro-ror	Solar-pv	Wind-Onshore	Battery	phs	Demand	Excess	phs-cos	Battery-cos
CONT	14.01	1.07	0.00	0.06	3.53	0.02	8.16	1.36	0.00	0.00	−28.0	−0.21	0.00	0.00
BASE	16.41	1.94	0.01	0.29	0.01	0.02	8.17	1.36	0.00	0.00	−28.0	−0.21	0.00	0.00
BASE-40	14.85	1.71	0.00	0.23	0.01	0.02	8.99	2.85	0.00	0.05	−28.0	−0.63	−0.08	0.00
BASE-50	12.73	1.13	0.00	0.13	0.00	0.02	9.14	6.37	0.00	0.14	−28.0	−1.45	−0.23	0.00
BASE-60	10.55	0.61	0.00	0.04	0.00	0.02	10.36	10.06	0.00	0.56	−28.0	−3.33	−0.88	0.00
BASE-70	8.05	0.33	0.00	0.02	0.00	0.02	11.37	16.69	0.15	0.72	−28.0	−7.99	−1.15	−0.21
BASE-80	5.58	0.02	0.00	0.00	0.00	0.02	14.33	18.50	2.18	0.69	−28.0	−9.30	−1.07	−2.95
BASE-90	2.80	0.00	0.00	0.00	0.00	0.02	19.03	18.44	4.80	0.63	−28.0	−10.25	−0.98	−6.49
GRE	19.44	0.00	0.20	0.00	0.00	0.00	8.54	0.00	0.00	0.00	−28.0	−0.19	0.00	0.00
GRE-40	16.60	0.00	0.20	0.00	0.00	0.00	9.48	2.38	0.00	0.20	−28.0	−0.55	−0.31	0.00
GRE-50	13.77	0.00	0.23	0.00	0.00	0.00	9.50	5.82	0.00	0.38	−28.0	−1.10	−0.60	0.00
GRE-60	10.90	0.00	0.30	0.00	0.00	0.00	10.38	10.08	0.00	0.57	−28.0	−3.35	−0.89	0.00
GRE-70	8.10	0.00	0.30	0.00	0.00	0.00	11.60	15.51	0.50	0.67	−28.0	−6.93	−1.06	−0.69
GRE-80	5.32	0.00	0.28	0.00	0.00	0.00	14.89	16.94	2.68	0.65	−28.0	−8.05	−1.03	−3.68
GRE-90	2.47	0.00	0.33	0.00	0.00	0.00	19.39	17.66	5.42	0.61	−28.0	−9.49	−0.98	−7.41
AUT	0.00	0.00	0.00	0.00	10.27	0.00	12.96	7.34	2.00	0.42	−28.0	−1.62	−0.66	−2.70
AUT-70	0.00	0.00	0.00	0.00	8.40	0.00	14.06	9.66	2.53	0.50	−28.0	−2.95	−0.78	−3.42
AUT-80	0.00	0.00	0.00	0.00	5.60	0.00	15.93	14.14	3.48	0.59	−28.0	−6.11	−0.92	−4.71
AUT-90	0.00	0.00	0.00	0.00	2.80	0.00	20.83	14.20	6.40	0.60	−28.0	−7.14	−0.96	−8.73

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Installed capacities in the four scenarios and the cost-optimal case in GW (left axis) and renewable energy (RE) share in percent (right axis).

Figure 2. Supply and demand in the four scenarios and the cost-optimal case in TWh (left axis) and CO2 emissions in million ton (right axis).

Figure 3. Installed capacities for all scenarios and varying renewable energy shares.

Figure 4. Supply/demand (left axis) and CO2 emissions (right axis) for all scenarios.

Figure 5. Dispatch of supply and demand in a week of the year for two different RE shares within the BASE scenario.

Figure 6. Aggregated SOC of pumped hydro storage (PHS) of BASE-40 (top) and BASE-80 (bottom) scenario.

Figure 7. LCOE for different scenarios and RE shares.

Table 1

Installed capacities in MW of Jordan’s electricity system in 2018 based on NEPCO data [7] and for 2023 based on [1] with planned projects and retirements of conventional units (CC: combined cycle gas turbine, GT: gas turbine, DE: diesel engine, ST: Steam turbine).

	CC	GT	ST	DE	Shale-ST	Wind	PV	Hydro
2018	2740	83	602	814	-	280.4	698.4	12
2023	2567	83	363	810	470	663	1144	12

Table 2

Scenario assumptions for the year 2030. Renewable energy profiles (FLH) have been calculated based on renewables.ninja [30,31]. For calculation of annualised investment, weighted cost of capital (WACC) 5% was applied in all scenarios.

	$\mathit{\eta }{}_{\mathit{u}}$	FOM	CAPEX	${\mathit{c}}_{\mathit{u}}^{\mathit{fuel}}$	FLH	Lifetime
	(-)	(%/${\mathit{c}}^{\mathit{capex}}$)	($/kW)	($/MWhth)	(h)	(Years)
Wind	1	3	1182 [32]	-	2050	20
PV	1	2	750 [32]	-	1912	20
CCGT	0.48 [1]	3.5	800 [1]	20.5	-	30
GT	0.33 [1]	3.5	550 [1]	20.5	-	30
ST	0.38 [1]	3.5	1300 [1]	20.5	-	30
DE	0.33 [1]	3.5		20.5	-	30
Oil shale ST	0.32 [1]	3	3720 [16]	25.2	-	30
Battery (power)	0.86 [33]	3	306 [33]	-	-	10
PHS (power)	0.80 [33]	1.5	1500 [34]	-	-	60
	(-)	(%/CAPEX)	($/kWh)	($/kWhth)	(h)
Battery (energy)	1	0	285 [33]	-	-	10