On-line version ISSN 2413-3051
Print version ISSN 1021-447X
J. energy South. Afr. vol.26 n.3 Cape Town Aug. 2015
Baraka KichongeI; Geoffrey R JohnII; Iddi S N MkilahaII
INelson Mandela African Institution of Science and Technology, Arusha, Tanzania
IICollege of Engineering and Technology, University of Dar es Salaam , Dar es Salaam, Tanzania
The current study applies an energy-system model to explore energy supply options in meeting Tanzania's electricity demands projection from 2010 to 2040. Three economic scenarios namely; business as usual (BAU), low economic consumption scenario (LEC) and high economic growth scenario (HEC) were developed for modelling purposes. Moreover, the study develops a dry weather scenario to explore how the country's electricity system would behave under dry weather conditions. The model results suggests: If projected final electricity demand increases as anticipated in BAU, LEC and HEC scenarios, the total installed capacity will expand at 9.05%, 8.46% and 9.8% respectively from the base value of 804.2MW. Correspondingly, the model results depict dominance of hydro, coal, natural gas and geothermal as least-cost energy supply options for electricity generation in all scenarios. The alternative dry weather scenario formulated to study electricity system behaviour under uncertain weather conditions suggested a shift of energy supply option to coal and natural gas (NG) dominance replacing hydro energy. The least cost optimization results further depict an insignificant contribution of renewable energy technologies in terms of solar thermal, wind and solar PV into the total generation shares. With that regard, the renewable energy penetration policy option (REPP), as an alternative scenario suggests the importance of policy options that favour renewable energy technologies inclusion in electricity generation. Sensitivity analysis on the discount rate to approximate the influence of discount rate on the future pattern of electricity generation capacity demonstrated that lower values favour wind and coal fired power plants, while higher values favour the NG technologies. Finally, the modelling results conclude the self-sufficiency of the country in generating future electricity using its own energy resources.
Keywords: electricity generation, MESSAGE model, discount rate, renewable energy, supply options
Energy is an important element in accomplishing the interrelated socio-economic development of any country. Tanzania's energy supply relies mainly on biomass, which accounts for nearly 90% of the total primary energy supply (Wawa, 2012, IEA, 2013). The remaining energy supply is accounted from petroleum products at approximately 8%, grid electricity 1% and renewable energy sources such solar and wind which account for nearly 1% (MEM, 2012, Kabaka and Gwang'ombe, 2007). Total electricity generation shares in 2012 were mainly from natural gas 50.7%, hydro 28.6%, oil products 20.1%, biofuels 0.3% and solar PV 0.2% (IEA, 2014). Projections are approximating electricity demand to reach 47.7 TWh in the year 2035 equivalent to an annual growth of approximately 8% (MEM, 2012). Energy resources are enormous and are available in various forms, including biomass, hydro, geothermal, biogas, wind, solar, natural gas and coal (Kihwele et al., 2012, MEM, 2013a). There is an estimated coal proven reserve of 304 million tonnes. whereas that of natural gas is 45 billion cubic meters (Kusekwa, 2013, MEM, 2012).
Geothermal has an estimated potential of 650 MW (Mnjokava, 2008, Kihwele et al., 2012), while hydro estimated potential is 4700 MW (MEM, 2012). Biomass estimated sustainable potential is 12 million TOE from agriculture wastes, plantation forests and natural forests (Wilson, 2010). The country experiences annual sunshine hours of 2 800 to 3500 hours and solar irradiation ranging from 4-7 kWh/m2 across the country (Casmiri, 2009, Kihwele et al., 2012). The renewable energy potential in the country is substantial but largely untapped for electricity and other thermal applications (Bauner et al., 2012, Kichonge et al., 2014b). The country's renewable energy potential from municipal solid wastes currently disposed in dump sites is considerable as shown in studies by Omari et al. (2014a) and Omari et al. (2014b).
Tanzanian energy demand, specifically electricity, has been growing over years because of socioeconomic transformations that opened up the country's economy. Statistics on the country's GDP growth since the year 2000 show an annual average increase of 7% (BOT, 2012). However, substantial challenges faces the electricity sector owing to constrained generation capacity and distribution network (Kapinga, 2013, Wangwe et al., 2014) which previously resulted in outages and rationing (Loisulie, 2010, MEM, 2013b). Despite the electricity sector challenges, the demand is expected to grow as the country targets a middle income economy status as detailed in Development Vision 2025 (URT, 1999) and its implementation through Big Results Now (BRN) initiatives (Kahyoza, 2013). Realizing Tanzania Development Vision 2025 goals implies that the country needs adequate, reliable, affordable and environmentally friendly electricity supply options. Achieving these require optimal generation capacity additions, which consider diversifications of power plants systems. Finding optimal generation capacity addition based on least cost plan is important in formulating supply options considering the high investments costs associated with it. It is therefore the objective of this study to apply MESSAGE (Model for Alternative Energy Supply Strategies and their General Environmental Impacts) to find least-cost optimal energy supply options. MESSAGE is an appropriate framework for this study as it is capable of dealing with long-term planning horizons based on high-resolution short-term system dynamics. Using MESSAGE, optimization of electricity supply options in each scenario will help describes possible future final electricity supply options availability. Study results will benefit policy and decision makers to arrive at a relevant solution interactively in national electricity system expansion planning.
2.1 MESSAGE Model
Model for Energy Supply Strategy Alternatives and their General Environmental Impacts (MESSAGE) is an optimization modelling tool (Messner and Strubegger, 1995), which calculates the least-cost energy supply system. Connolly et al. (2010), describes MESSAGE as a bottom-up model capable of optimizing operation and investment of technologies in medium to long term energy systems. MESSAGE modelling approach allows the realistic evaluation of the long-term role of an energy supply option under competitive conditions (IAEA, 2008, Hainoun et al., 2010). The least-cost determination in MESSAGE is through minimization of the total discounted energy system cost subject to the constraints representing demands, resource deficiency and capacity limits. Discounted energy system cost minimization includes investments, fixed and variable operation costs, maintenance costs, fuel and any additional penalty costs, which defines the limits and constraints relation. With MESSAGE, alternative energy supply strategies in agreement with user-defined constraints are assessed (IAEA, 2006, Tait et al., 2014).
Mathematical techniques tied up with MESSAGE comprises of linear and mixed-integer programming. The purpose for linear programming (LP) applications is that all the limits and the objective function (optimization target) are linear functions of the decision variables. Mixed-integer use in MESSAGE is due to integer values at the optimal solution requirements by some of the decision variables. Objective function in MESSAGE modelling approach is as shown in Equation 1. The variable Xi,j,tdenotes a flow variable (input) of fuel form i in technology j in the time step t. Flow variable describes amount produced in which technology and the type of fuel. The investment variable denoted by Yi,t represents new installation of technology j in time step t.
The MESSAGE model computes the objective function to satisfy the condition to ensure demand-supply balance as illustrated in equation 2. The parameter D denotes energy demand, j represents energy demand of j while t represents time step. In addition, n represents technology efficiency, X denotes production decision of the technology, i is the number of technologies and n total number of technologies.
MESSAGE has been used to model the power supply sector by means of the principle of reference energy system (RES), which allows representation of the entire energy network including possible development paths (Recka, 2011, Selvakkumaran and Limmeechokchai, 2011). RES is composed of energy resources and sources, energy carriers (form) and technologies. RES captures network flow of energy carrier from one process to the other starting in the resource to the consumer delivery. The explanation of energy forms includes each level of energy chains, technologies using or producing these energy forms, and the energy resources. MESSAGE defines energy forms and technologies in all steps of energy chains. This includes identification of energy chain levels beginning from the demand to the resources, the energy forms to energy services. MESSAGE computes energy demand from the first level of each energy chain up to the energy resource level. Final demand level is distributed according to the types of consumption (Van Beeck, 1999, Pinthong and Wongsapai, 2009).
The MESSAGE modelling approach has previously applied to formulate an optimal energy supply strategy for Syria (Hainoun et al., 2010); policy options study for power sector in Zambia (Tembo, 2012); strengthening of renewable energy applications (IAEA, 2007; IAEA, 2006); Optimal electricity system planning in a large hydro jurisdiction: Will British Columbia soon become a major importer of electricity? (Kiani et al., 2013) alternate electricity supply model (Roque, 2014); climate change policy analysis (Nakicenovic et al., 2000): nuclear energy in mitigating CO2emissions (AlFarra and Abu-Hijleh, 2012) among many others. Further information on MESSAGE as LP optimization tool is as found at the IAEA organization web site (www.iiasa.ac.at/web/home/research/modelsData/MESSAGE/MESSAGE.en.html).
2.2 Electricity demand projections
The final electricity demand projections were done using Model for Analysis of Energy Demand (MAED) (Kichonge et al., 2014a) and have been summarized in Figure 1. MAED is a bottom-up modelling approach (Bhattacharyya and Timilsina, 2009) chosen because of its suitability to model the final electricity demand projections based on time and data availability. Suitability of MAED to relate systematically the corresponding social, technological and economic factors which affect the demand was also considered in the selection of the model (IAEA, 2006, IAEA, 2009). Literatures such as Hainoun et al. (2006), IAEA (2006) ), IAEA (2009), Nakarmi et al. (2013) and the IAEA organization website (www.-pub.iaea.org/MTCD/publications/PDF/CMS-18_web.pdf ) gives detailed account of MAED.
2.3 Modelling framework
2.3.1 Electricity conversion technologies Conversion technologies considered includes coal fired power plant, solar PV, hydro, solar thermal, biomass, conventional gas turbine (GT), heavy fuel oil (HFO) and combined cycle gas turbine (CCGT) power plants.
2.2.2 Reference energy system (RES)
The proposed Tanzanian RES accommodates resources, primary, secondary and final demand energy levels. Simplified schematic flow of the energy chains, levels and conversion technologies in RES are as described in Figure 1. Rectangles in the RES represent the technologies, which contain the techno-economic data. A single technology as used in the proposed RES denotes all conversion technologies, which uses the same type of fuel. The energy resource level is characterized by coal and natural gas, which are locally available resources. Energy carriers in the form of natural gas (NG), coal and HFO defines primary energy level in the energy chain. The secondary energy level is composed of electricity as the only form of energy echoed in this study. Intermediary of primary and secondary energy levels, there are electricity conversion technologies whose main inputs are energy carriers from the primary energy level. Electricity transmission and the distribution network connect secondary and final energy levels. The final electricity demand developed from the model's external factors (Kichonge et al., 2014a) is given at the first level of each energy chain. The model calculates the equivalent productions of each of the technologies at the succeeding levels of the chain up to the energy resource level, which then gives the optimal technical choice by minimizing the total system cost, while meeting the given final electricity demand.
2.4 Modelling basic assumptions
The general assumptions considered in modelling energy supply options for electricity generation for Tanzania are as follows:
- All model scenarios span from 2010, which is the base year to 2040 as the last year. A time step of five years has been adopted throughout the study period as more steps slows down the solver and also for easy results reporting;
- Each model year in all scenarios is divided into four seasons to capture seasonal variations in reservoir inflows and load for hydro, solar PV and wind turbines. The seasons includes Season 1, which encompasses January to February (dry season); Season 2 - March to May (wet/rainy season); Season 3 - June to September (dry/sunny weather season) and Season 4 -October to December (short rainy season);
- The expected load profile for defining the mix in power generation plants follows an annual hourly and monthly load curve characteristics as shown in Figures 3 and 4. An annual hourly load curve with characteristics was produced from hourly generation data collected for the years 2009 to 2012. Generation of annual hourly load curves was done by taking average values in load demands for a particular hour throughout a year. Daily base load patterns together with energy resources variations are taken into account by describing two types of days which are workdays (Monday to Saturday) and weekends (Sunday and holidays). The daily base load patterns for a 24 hour day has been divided as nine parts for Season 1, ten parts for Season 2, eight parts for Season 3 and twelve parts for Season 4;
- Final electricity demand differences under business as usual (BAU), low economic consumption (LEC) and high economic growth (HEC) scenarios as projected in MAED are as depicted in Figure 1. Other parameters such as energy forms, seasonal and daily power demand variability, constraints, technologies and resources remained the same for BAU, LEC and HEC scenarios.
- Air emissions control measures have not been included in the model;
- The operation time thus electricity output for solar PV, solar thermal and wind power plants follows the proposed seasonal and daily sunshine/wind variation;
- Geothermal power plants begin operation in 2025, with an initial installed capacity of 100 MW and increasingly to 650 MW in 2040 (MEM, 2012, Mnzava and Mayo, 2010);
- The discount rate parameter for economic evaluation of the future investment project was set to 10% in each scenario;
- The entire national electricity system has been simplified and modelled as a single grid system;
- Existing and future expansion projects, transmission and distribution losses and reserve margins as specified in the power system master plan (MEM, 2012) has been adopted for optimization purposes;
- Summary of the crucial parameters for modelling electricity supply options in terms of specific technical and economic characteristics adopted for conversion technologies are as depicted in Table 1.
- The investment costs for renewable energy technologies (wind, solar PV and thermal) assumed a decreasing trend as the industry develops and thus becomes cost competitive in future (Philibert, 2014). The investment cost for wind technology in the base year as shown in Figure 5 was approximated at 2 438 US$/kW and then decreased steadily to 1 800 US$/kW in 2025 where it assumed this constant value to 2040. Solar PV technology investment costs assumed a base year value of 4 000 US$/kW and decreased in steps to 3 500 US$/kW and 2 500 US$/kW in 2025 and 2030 respectively where it presumed a constant value of 2 500 US$/kW towards the year 2040. Similarly, the investment cost for solar thermal technology towards the end of the study period presumed a decreasing trend from the base year value of 4 500 US$/kW to 3 500 US$/kW.
3. Results and discussions
Final electricity demands have been optimized in order to determine the optimal energy supply options for Tanzanian electricity sector. This section presents MESSAGE modelling results calculated based on the least-cost energy supply options for electricity generation for the period 2010-2040. Based on the total system costs of the electricity system discounted over the study period 2010-2040, three different energy supply options have been optimized in lieu of BAU, LEC and HEC scenarios as detailed in Kichonge et al. (2014a).
3.1 Installed capacity
The total installed capacity increases gradually from 804.2 MW in the base year to 10 811.5 MW, 9 190.6 MW and 13 325.6 MW in 2040 for BAU, LEC and HEC scenarios respectively as illustrated in Figure 6. The least-cost optimal results show HEC scenario has the highest total capacity additions at 12 521.4 MW in 2040 as compared to the BAU scenario 10 007.3 MW and LEC scenario 8 386.65 MW. Annual increase of installed capacity in HEC scenario is equivalent to 9.81%, while BAU and LEC scenarios projection increases are 9.05% and 8.46% respectively. Hydro, NG, coal and geothermal power plants dominate the total installed capacity additions in all scenarios. Wind and biomass represents a small proportion in the total installed capacity whereas solar PV and thermal were not able to compete.
There is a corresponding increase of thermal installed capacity addition (coal and NG power plants) in both scenarios. NG power plants (CCGT and GT) increase their shares in the total installed capacity from 202 MW in 2010 to 2 546.55 MW in 2040 for BAU scenario. LEC scenario observes a similar increasing trend to 2 090.67 MW in 2040, while HEC scenario is 4794.47 MW. The shares of hydro power plants witnesses an opposite decreasing trend in the period 2015-2030, where it picks-up the dominance to 2035. Hydro power plants shares decrease from 69.8% in 2010 to 35.8%, 42.1% and 29% in 2040 for BAU, LEC and HEC scenarios respectively. The main reason attributed to the decreasing trend is the potential constraints despite the fact that it is the cheapest in operating costs (MEM, 2013b).
3.2 Least cost electricity generation mix
Summarized least cost total electricity generation for each scenario are as shown in Figure 7 and the least-cost electricity generation supply options results by technology in Table 3. BAU scenario least cost electricity generation expanded from 5 632 GWh in 2010 to 62 770 GWh in 2040. The expansion is equivalent to an annual growth rate of 8.4% as compared to 7.9% and 9.3% for the LEC and HEC scenarios respectively. The base year proportions in the generation mix include hydro (66.7%), NG (28.9%), biomass (2.5%) and HFO (2%).
Results describe general dominance of hydro power plants in the generation mix with NG, biomass and HFO power plants compensating the balance. The optimized results show the proportion of hydropower plants generation increasing gradually to 41.2 %, 47.1 % and 31.7 % in 2040 for BAU, LEC and HEC scenarios respectively.
The proportion of coal power plants in the total generation rises gradually from 11.8 % in 2020 to 38.1% in 2040 for BAU scenario while for LEC scenario is 9.3% in 2020 and rises to 33.5% in 2040.
HEC scenario witness higher proportion at 17.7% in 2020 and grows to 33% in 2040. The higher proportion of hydro, NG and coal power plants in the generation mix is imminent due to lower investment and fuel costs as compared to other candidates technologies considered. Unlike the increases in hydro and coal power plants generation shares, the proportions of NG power plants increasingly dominate more than 50% of the generation from 2015-2020 and thereafter declines in 2040. The decline in NG power plants proportions after 2020 is due to presumed new investments in hydro power plants. NG power plants technologies proportion in 2040 for BAU scenario split into CCGT (96.9%) and GT (3.1%). However, a similar trend follows in LEC and HEC scenarios in which the share of CCGT will be as high as 95.7% and 99.3% respectively in 2040. The choice of CCGT in least-cost optimized results attributes to higher availability and efficiency up to 60% in comparison to that of GT 40% (Sharman, 2005, Sims et al., 2003). Combined thermal generation contributes 48.8% of the total in 2040 for BAU scenario whereas the contribution of thermal generation for LEC scenario is 44.1 % and that for HEC scenario is 60.6%. Higher final electricity demands in HEC scenario drives the increase in the use of thermal generations. Hydro and geot-hermal resource potential constraints attributes to the use of more thermal generations towards the end of the study period. On the contrary, renewable energy with the exclusion of hydro makes up small proportion in the contribution of the total electricity generation mix.
The contribution of renewables technologies into electricity generation for BAU, LEC and HEC scenarios extends to 6.2 TWh, 4.8 TWh and 6.2 GWh respectively in the year 2040. The share of renewable energy generation in BAU scenario accounted for an average of 2.1% in the period from 2010 to 2025 and, thereafter, grows to 15.4% in 2030 and then retreats to 9.9% in 2040. The rise of renewable energy in 2030 attributes to utilization of full geothermal energy potential presumed in the year. HEC scenario shares of renewable energy from 2025 - 2040 averaged at 1.9 % in the total electricity generation. Comparable trends are also as observed in LEC scenario. Moreover, within renewable energy technologies, geothermal dominates the generation mix followed by biomass and wind with insignificant shares from solar thermal and solar PV power plants. Geothermal and wind power plants by the end of the study period in 2040 generated 7.7 % and 2.2 % respectively of all electricity in the BAU scenario. Similarly, geothermal power plants generation for LEC and HEC scenarios was approximately 8.8 % and 5.9 % respectively. The constraints on geothermal energy resources potential and the rise in electricity demand reduced the share of geothermal technologies in 2040 for HEC scenario.
The least-cost electricity generation results in the BAU, LEC and HEC scenarios draws four most important conclusions. The first one is the key role played by hydro, coal and NG technologies in the final electricity generations. These technologies have shown least-cost competitiveness in electricity generation, which describes their importance in sustainable development of the electricity sector. The second is insignificant contribution from solar thermal and solar PV technologies in the entire study period. The high investment costs associated with the technologies discourages the penetration in generations mix despite their least operations and maintenance costs. The last conclusion is the country self-sufficiency in generating electricity using its own local energy resources thus ensuring security of supply for sustainable development.
3.3 Primary energy supply
Primary energy supply composition for electricity generation is as shown in Table 4. Coal, NG, HFO and biomass are the main primary energy supply for electricity generation. Conversion technologies for geothermal, hydro, wind, solar PV and solar thermal do not consume primary energy for electricity generation. Primary energy supply in the BAU scenario will grow from 6 203 GWh in 2010 to 73 083 GWh in 2040. Similarly, the growth in the LEC scenario amounts to 57 529 GWh against 110,700 GWh in the HEC scenario. Generally, all scenarios projects increased coal consumptions as compared to NG with small proportions from biomass and HFO towards 2040. The least-cost supply option, show electricity generation will depend on coal and NG to cover primary energy supply. It further depicts gradually decrease in HFO to less than 0.1% in 2040. Figure 8 depicts primary energy supply in the BAU scenario that is the representative trend for other scenarios.
3.4 CO2 emissions
The total CO2 emission depicted in Figure 9 rises from 1182-kilo tonnes of CO2 in 2010 to 23652.3-kilo tonnes of CO2 in 2040 for the BAU scenario. The rise in CO2 emission in the BAU scenario represents an annual growth of 10.5%. Similarly, the increases for LEC and HEC scenarios represent annual growth of 9.5% and 11.7 % respectively. The growth rate in CO2 emissions in the LEC scenario is much lower than that of HEC and BAU scenarios due to slow economic growth presumed in the scenario representing less energy consumption. Similarly, the higher CO2 emissions in the HEC scenario is highly influenced by higher electricity demands, which resulted in optimal capacity additions of coal and NG power plants. The emission of CO2 in all scenarios is higher due to insignificant renewable energy conversion technologies applications.
3.5 Economics of scenarios
The capital investment cost required for the entire period of study is based on MESSAGE least cost modelling results. The sharing of capital investments for the period of 2010 - 2045 is as shown in Figure 10. In meeting final electricity demand under BAU, LEC and HEC scenarios, the total capital investment cost of 4 488 million US$, 3 903 million US$ and 5 573 million US$ respectively would be required. The main share of the capital investments for the entire period in BAU, LEC and HEC scenarios falls into the period of 2015 to 2035, in which most of the capacity addition is taking place. The capital investment needed to develop a BAU scenario final electricity demand in the entire study period would be about 584 million US$ more than LEC scenario, while 1 086 million US$ increase would be needed for a HEC scenario. The higher capital investment costs are observed in 2035 for HEC and BAU scenarios, while for LEC scenario is in 2030.
Least cost modelling results as presented in Table 5 shows the main differences among BAU, LEC and HEC scenarios in terms of the investment cost and variable and fixed O&M costs. Variable O&M costs for the HEC scenario are 335.1 and 529.4 million US$ higher than those for the BAU and LEC scenarios. Moreover, the LEC scenario entail lower fixed O&M costs at 1013.4 million US$ as compared with the BAU and HEC scenarios.
4. Sensitivity analysis
Sensitivity analyses carried out in this study, intended to explore the influence of techno-economic parameters, policy options and extreme dry weather conditions in the expansion of the final electricity generation mix.
4.1 Renewable energy penetration
Least-cost optimization results as distinguished in previous sections reveals in-significant penetration into electricity generation of renewable energy technologies due to the high investment costs. Among the reasons behind the insignificant penetration of renewable energy technologies in electricity generation is the absence of non-environmental friendly energy supply constraints. As a result, the market forces decide to choose the least-cost energy supply options for electricity generations, which in most cases, occurs as non-environmentally friendly sources (Bull, 2001; Lewis, 2007). Based on this fact, the study formulates a renewable energy penetration policy option (REPP) as an alternative scenario to study electricity system behaviour under energy supply constraints to promote renewable energy technologies in the generation of electricity. The policy option in REPP requires a compulsory penetration of renewable energy technologies (combined together) to contribute at least 10% of the total electricity generation in 2020 and increasingly to 30% in 2040. The REPP scenario assumes energy demands projections and all techno-economic parameters of the BAU scenario with the additions of the compulsory policy measures. All modelling inputs of REPP scenario remain the same as in the BAU scenario except for the imposed compulsory penetration of renewable energy technologies.
The results of REPP scenario implementations as compared to the BAU scenario in terms of the total installed capacity, electricity generation and CO2 emissions are as illustrated in Table 6. The MESSAGE results depict a huge reduction of CO2 emissions at approximately 48% in the REPP scenario as compared to the BAU scenario in 2040. The displacement of thermal power plants with renewable energy technologies has resulted in reduction of CO2 emissions and primary energy supply. The total installed capacity shares of renewable energy technologies increases to 17.1% and 34.7% in 2020 and 2040 respectively. BAU scenario composition was 0.8% in 2020 and 10.6% in 2040. The shares of renewable energy technologies in the total generation mix for the REPP scenario has increased to 30% in 2040 as compared to 9.9% it had in the BAU scenario. Satisfactory inclusion of renewable energy technologies into the electricity generation mix, as shown in the REPP scenario, has demonstrated the importance of compulsory measures in policy formulation in favour of renewables.
Even though the compulsory policy measures resulted in the expansion of renewable energy technologies shares, REPP scenario depict additional investments costs as compared to the BAU scenario. The comparison in the investments costs between the BAU and REPP scenarios is as depicted in Figure 11. Meeting REPP scenario requirements will necessitate considerable investment cost of 7 665.8 million US$ as compared to 4 487.6 million US$ for the BAU scenario. Contrariwise, as shown in Figure 12, REPP scenario accommodation exhibits a decrease in the operation and maintenance variable costs (O&M). There is a decrease to 680.6 million US$ in the operation and maintenance variable costs for the REPP scenario, when compared to 999.7 million US$ for the BAU scenario in the entire study period. MESSAGE modelling results; show that the REPP scenario demands a more aggressive approach to investment in renewable energy technologies. For that reason, if the country chooses to implement the policy, additional policies such as renewable energy feed-in tariff and institutional frameworks that are essential for the growth of renewable energy technologies must be in place. The compulsory policy measures as revealed in MESSAGE modelling helps in tapping of the enormous potential of renewable energy resources into electricity generations for the benefit of the environment and security of supply.
4.2 Discount rate adjustments
Adjustments were carried out on the BAU scenario to approximate the influence of the discount rate on the upcoming pattern of electricity generation capacity, electricity production or economic effectiveness of a number of electricity generation plants. The discount rate adjustments carried out were 6%, 8%, 12% and 14% in comparison to a study-adopted value of 10%. Adjustments of discount rate value to 6% preferred early capacity addition of 500 MW wind power plant into electricity generation in 2035, while 8%, 10% and 12% favours addition in 2040 with no addition for the 14% values. Solar PV and thermal failed to be competitive in all discount rates adjustments. These technologies require special policy option for their inclusion into electricity generation to be realistic. Adjustments of discount rate values to 12% and 14% were in favour of capacity addition of CCGT and GT power plants as opposed to lower values (6% and 8%), which preferred coal power plants. Higher efficiency coupled by a lower operating and maintenance costs, shorter construction time and fuel cost characterizes CCGT and GT power plants thus turn out to be more attractive for capacity addition in comparison to other technologies options. The discounts rates of 6%, 8%, 12% and 14% resulted in coal fired power plants total installed capacity of 6 652 MW, 6 108 MW, 5 385 MW and 4773 MW respectively. In other words, a higher value of discount rate leads to the postponement of large-scale investments. According to the minimum cost criterion, a discount rate of 10% gives greater preference for the fossil fuel scenarios. A decrease or increase of the discount rate has insignificant influence on capital investments of hydro, biomass and geother-mal, which seems to be due mainly to the limited resources potential.
4.3 Dry weather scenario
Experience has shown weather conditions affect electricity generation capacity causing outages and rationing (Loisulie, 2010; MEM, 2013b). The alternative dry weather scenario was formulated to analyse electricity system behaviour under uncertain weather conditions. All modelling inputs of a dry weather scenario remain the same as in the BAU scenario except for the imposed generation's constraints of hydropower to 20% of the total generations in the period 2020-2040. The least-cost results as shown in Table 7 shows the generations will shift to coal and NG power plants at approximately 42.8% and 30.2% respectively. The capacity additions for coal power plants will expand to 9 772 MW as compared to 6 040 MW in the BAU scenario. Because of imposed hydropower constraints, the CO2 emission will increase to 86938.67-kilo tonnes of CO2 as compared to 51296.6-kilo tonnes of CO2in the BAU scenario. Based on MESSAGE modelling results, if the country chooses to implement measures because of dry weather conditions, more usage of coal and NG as primary energy supplies will be the least cost option. The additional capacity in terms of coal and NG power plants to replace hydropower plants would decrease both the risks of a dry weather condition and energy security uncertainties. However, the weaknesses of coal and NG development into a dry weather scenario are the higher CO2 emissions as compared to the BAU scenario as depicted in Figure 13. The capital investment cost of the dry weather scenario will require less than 535.52 million US$ as compared to capital investment in the BAU scenario. Less capital investment cost in the dry weather scenario is due to lower capital investment cost and shorter construction time of coal coal-fired and NG power plants. Despite hydro power plant lower operation and maintenance costs, coupled with zero fuel consumption for final electricity generation, they have higher capital investment costs and longer construction life (Sharma, 2010).
The study presented a modelling approach on the energy supply options for electricity generation in Tanzania. The modelling approach emphasized optimal results based on the least-cost as assumed in MESSAGE. Based on the results presented, MESSAGE turned out to be a useful tool to address energy supply options for electricity generation in Tanzania. The projected total installed capacity increases gradually from 804.2 MW in the base year to 10811 MW, 9190.9 MW and 13325.6 MW in 2040 for the BAU, LEC and HEC scenarios respectively. The increase in the total installed capacity would call for capital investment cost of 4 488 million US$, 3 903 million US$ and 5 573 million US$ respectively for the BAU, LEC and HEC scenarios.
Hydropower plants dominate the capacity additions followed by coal, CCGT, geothermal and GT power plants to meet the electricity generation expansion in both scenarios. Total primary energy supply dominated by coal and NG rises to 73 083 GWh, 57 529 GWh and 110 700 GWh in 2040 for the BAU, LEC and HEC scenarios respectively, as compared to base year amount of 6 203 GWh. In meeting final electricity demands, CO2 emissions will expand from 1182-kilo tonnes of CO2 to 10.5%, 9.5% and 11.7% respectively for BAU, LEC and HEC scenarios with decreases of CO2 in the REPP scenario.
Renewable energy sources as concluded in the REPP scenario were identified as promising for meeting the future electricity demand in Tanzania.
Potential contribution of renewable energy sources to the savings of coal and NG reserves would be a great contribution to the economy and the environment. However, the dry weather scenario has shown a shift to coal and NG power plants generations at approximately 42.8% and 30.2% respectively resulting into higher CO2. The sensitivity analysis tests results have shown lower discount rates to favour investments on wind and coal power plants, while higher discount rates favour NG power plants. The least-cost results have shown implications concerning capital investment costs versus environmental impacts concerns. Least cost modelling results have concluded that meeting final electricity demands without considerations of environmental impacts concerns is cheaper. Policy makers should balance the capital investment costs and environmental concerns in the energy planning of the country.
The authors acknowledge the Tanzania Atomic Energy Commission (TAEC), Arusha Technical College (ATC), Nelson Mandela African Institution of Science and Technology (NM-AIST) and The College of Engineering Technology (CoET) of the University of Dar es Salaam (UDSM) for their kind support in successful completion of this study. Finally yet importantly, the authors wish to acknowledge valuable a contribution from Sadiki Mutabuzi of TAEC.
AlFarra, H. J. & Abu-Hijleh, B. (2012). The potential role of nuclear energy in mitigating CO2 emissions in the United Arab Emirates. Energy Policy, 42, 272-285. [ Links ]
Bauner, D., Sundell, M., Senyagawa, J. & Doyle, J. (2012). Sustainable Energy Markets in Tanzania Report II: Analysis and conclusions [Online]. [ Links ] Swedish Agency for International Development Cooperation (SIDA) and the British Department for International Development (DfID). Available: http://www.renetech.net/wp-content/uploads/2013/03/Sustainable_Energy_Markets_in_Tanzania_II_final.pdf [Accessed 28th June 2013].
Bhattacharyya, S. C. & Timilsina, G. R. (2009). Energy demand models for policy formulation: a comparative study of energy demand models. [ Links ]
BOT (2012). Economic Bulletin for the Quarter Ending June, 2012 Dar es Salaam: Bank of Tanzania. [ Links ]
Bull, S. R. (2001). Renewable energy today and tomorrow. Proceedings of the IEEE, 89, 1216-1226. [ Links ]
Casmiri, D. (2009). Energy Sytems: Vulnerability -Adaptation - Resilience (VAR) [Online]. [ Links ] HELIO International. Available: http://www.helio-international.org/VARTanzania.En.pdf [Accessed 12th February 2013].
Connolly, D., Lund, H., Mathiesen, B. V. & Leahy, M. (2010). A review of computer tools for analysing the integration of renewable energy into various energy systems. Applied Energy, 87, 1059-1082. [ Links ]
Hainoun, A., Seif-Eldin, M. & Almoustafa, S. (2006). Analysis of the Syrian long-term energy and electricity demand projection using the end-use methodology. Energy Policy, 34, 1958-1970. [ Links ]
Hainoun, A., Seif Aldin, M. & Almoustafa, S. (2010). Formulating an optimal long-term energy supply strategy for Syria using MESSAGE model. Energy policy, 38, 1701-1714. [ Links ]
IAEA. (2006). Assessing Policy Options for Increasing the Use of Renewable Energy for Sustainable Development: Modelling Energy Scenarios for Ghana [Online]. [ Links ] Vienna, Austria: International atomic Energy Agency. Available: www.unep.org/training/programmes/.../Part_2/.../Energy/.../Ghana.pdf.
IAEA. (2007). Assessing policy options for increasing the use of renewable energy for sustainable development: Modelling energy scenarios for Sichuan, China [Online]. [ Links ] Vienna, Austria: International Atomic Energy Agency. Available: http://www.iaea.org/OurWork/ST/NE/Pess/assets/sichuan_case_study.pdf.
IAEA (2008). Guidance for the Application of an Assessment Methodology for Innovative Nuclear Energy Systems. INPRO Manual-Overview of the Methodology International Atomic Energy Agency. (IAEA-TECDOC-1575 Rev. 1): Vienna. [ Links ].
IAEA. (2009). IAEA Tools and Methodologies for Energy System Planning and Nuclear Energy System Assessments [Online]. [ Links ] Vienna, Austria: International Atomic Energy Agency. [Accessed 16th October 2012].
IAEA. (2006). Model for Analysis of Energy Demand (MAED-2) [Online]. [ Links ] Vienna, Austria: International Atomic Energy Agency. Available: http://www-pub.iaea.org/MTCD/publications/PDF/CMS-18_web.pdf [Accessed 12th December 2012].
IEA. (2013). Energy balances of non-OECD countries 2013 [Online]. [ Links ] Paris, France: International Energy Agency. Available: http://www.iea.org/statistics/statisticssearch/report/?&country=TANZANIA&year=2011&product=Indicators.
IEA. (2014). Energy Balances of non-OECD Countries 2014 [Online]. [ Links ] Paris, France: International Energy Agency. Available: /content/book/energy_bal_non-oecd-2014-en http://dx.doi.org/10.1787/energy_bal_non-oecd-2014-en.
Kabaka, K. T. & Gwang'ombe, F (2007). Challenges in small hydropower development in Tanzania: rural electrification perspective. International Conference on Small Hydropower-Hydro. Sri Lanka, 2007. 22-24. [ Links ]
Kahyoza, N. (2013). Tanzania Gas Sector Economy in the Light of Human Development [Online]. [ Links ] Dar es Salaam: Economic and Scocial Research Foundation (ESRF). Available: http://esrf.or.tz/docs/EPB-5-GasSector_HumanDev.pdf No. 5/2013].
Kapinga, S. (2013). A study on sustainable electricity: The Use of Energy Mix in Modernizing Tanzania. [ Links ]
Kiani, B., Rowe, A., Wild, P, Pitt, L., Sopinka, A. & Pedersen, T. F. (2013). Optimal electricity system planning in a large hydro jurisdiction: Will British Columbia soon become a major importer of electricity? Energy Policy, 54, 311-319. [ Links ]
Kichonge, B., John, G. R., Mkilaha, I. S. & Sameer, H. (2014a). Modelling of Future Energy Demand for Tanzania. Journal of Energy Technologies and Policy, Vol. 4, Issue No 7. [ Links ]
Kichonge, B., Mkilaha, I. S. & John, G. R. (2014b). Prediction of the Contribution of Renewable Energy Sources into Electricity Generation in Tanzania. International Journal of Renewable Energy Technology Research (IJRETR), Vol. 3 1 - 13. [ Links ]
Kihwele, S., Hur, K. & Kyaruzi, A. (2012). Visions, Scenarios and Action Plans Towards Next Generation Tanzania Power System. Energies, 5, 3908-3927. [ Links ]
Kusekwa, M. A. (2013). Biomass Conversion to Energy in Tanzania: A Critique. [ Links ]
Lewis, N. S. (2007). Toward cost-effective solar energy use. science, 315, 798-801. [ Links ]
Loisulie, S. (2010). Vulnerability of the Tanzanian Hydropower Production to Extreme Weather Events [Online]. [ Links ] Morogoro: Sokoine University of Agriculture Faculty of Science. Available: http://cdsagenda5.ictp.Meste.it/html_trees_links/1201801804/participants_presentations/Loisulie-Tanzania_hydropower.pdf [Accessed 4th May 2010].
MEM (2012). Power System Master Plan 2012 Update. Ministry of Energy and Minerals-Dar es Salaam. [ Links ]
MEM (2013a). The Energy Sector Overview [Online]. [ Links ] Ministry of Energy and Mineral-Tanzania. Available: http://www.mem.go.tz/AboutUs/MEMOverview/TheEnergySector/tabid/121/language/en-US/Default.aspx [Accessed 10th January 2013].
MEM (2013b). Scaling Up Renewable Energy Programme (SREP) - Investment Plan for Tanzania [Online]. [ Links ] Ministry of Energy and Minerals - United Republic of Tanzania. Available: http://www.mem.go.tz/Portals/0/EasyDNNNewsDocuments/1033/0039_08052013-SREP_Tanzania%20IP_Final_Draft.pdf [Accessed 12th July 2013].
Messner, S. & Strubegger, M. (1995). User's Guide for MESSAGE III. Rep. WP-95-69, International Institute for Applied Systems Analysis, Laxenburg, Austria, 14. [ Links ]
Mnjokava, T. (2008). Geothermal exploration in Tanzania-Status report. Presentation by Geological. [ Links ]
Mnzava, L. J. & Mayo, A. W. (2010). Geochemical investigation of geothermal power potential exploration of hot springs in South-western Tanzania. ABOUT IJWREE, 601. [ Links ]
Nakarmi, A. M., Mishra, T. & Banerjee, R. Energy Scenarios of Nepal 2010-2030: A Renewable and Sustainable Approach. Proceedings of the International Conference on Energy Resources and Technologies for Sustainable Development (ICERTSD), 2013 Howrah, West Bengal, India [ Links ]
Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J & GafQn S. (2000). Special report on emissions scenarios. Intergovernmental panel on climate change [Online]. [ Links ] Available: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.260.1482.
Omari, A., Kichonge, B., John, G. R., Njau, K. N. & Mtui, P L. (2014a). Potential of municipal solid waste, as renewable energy source - a case study of Arusha, Tanzania. International Journal of Renewable Energy Technology Research (IJRETR), Vol. 3,No. 6, pp. 1 - 9. [ Links ]
Omari, A., Said, M., Njau, K., John, G. & Mtui, P. (2014b.) Energy Recovery routes from Municipal Solid Waste, A case study of Arusha-Tanzania. Journal of Energy Technologies and Policy, 4, 1-7. [ Links ]
Philibert, C. (2014). Technology roadmap: solar photovoltaic energy, OECD/IEA. [ Links ]
Pinthong, T. & Wongsapai, W. Evaluation of Energy Demand and Supply Under Electricity Generation Scenarios of Thailand. Proceedings of the World Renewable Energy Congress, 2009. [ Links ]
Recka, L. (2011). Electricity system optimization: A case of the Czech electricity system - application of model MESSAGE. [ Links ]
Roque, P. M. (2014). Mozambique's electricity demand forecasting and power supply alternative model. The 9th Regional Collaboration Conference -Research and Innovations forum for sustainable regional Development: Over a Decade of Swedish Partnership, July 20 - 23, 2014 2014 Imperial Resort Beach Hotel, Entebbe, Uganda. [ Links ]
Selvakkumaran, S. & Limmeechokchai, B. Assessment of Thailand's energy policies on energy security. Utility Exhibition on Power and Energy Systems: Issues & Prospects for Asia (ICUE), 2011 International Conference and, 2011. IEEE, 1-6. [ Links ]
Sharma, A. (2010). Hydro power vs thermal power: a comparative cost-benefit analysis. International Journal of Arts and Sciences, 3, 125-143. [ Links ]
Sharman, H. Why UK wind power should not exceed 10 GW. Proceedings of the ICE-Civil Engineering, 2005. Thomas Telford, 161-169. [ Links ]
Sims, R. E., Rogner, H.-H. & Gregory, K. (2003). Carbon emission and mitigation cost comparisons between fossil fuel, nuclear and renewable energy resources for electricity generation. Energy policy, 31, 1315-1326. [ Links ]
Tait, L., McCall, B. & Stone, A. (2014). Energy futures modelling for African cities-selecting a modelling tool for the SAMSET project. [ Links ]
Tembo, B. (2012). Policy Options for the Sustainable Development of the Power Sector in Zambia. M Sc Thesis, University of Cape Town. [ Links ][Links].
URT (1999). The Tanzania Development Vision 2025. Dar es Salaam: President's Office. [ Links ]
Van Beeck, N. 1999. Classification of energy models, Citeseer. [ Links ]
Wangwe, S., Mmari, D., Aikaeli, J., Rutatina, N., Mboghoina, T. & Kinyondo, A. (2014). The performance of the manufacturing sector in Tanzania: Challenges and the way forward. WIDER Working Paper. [ Links ]
Wawa, A. I. (2012). The challenges of promoting and adopting biogas technology as alternative energy source in semi-arid areas of Tanzania: The case of Kongwa and Bahi districts of Dodoma region. PhD, The Open University Of Tanzania [ Links ]
Wilson, L. (2010). Biomass Energy Systems and Resources in Tropical Tanzania. Royal Institute of Technology; KTH: ISBN 978-91-7415-732-1 [ Links ]
Received 3 November 2014
29 July 2015