Techno-Economic Comparison of Microgrids and Traditional Grid Expansion: A Case Study of Myanmar

Abstract

Myanmar’s electricity supply relies mainly on hydropower and gas-fired generation, yet rural electrification remains limited, with national access at approximately 60%. The National Electrification Plan (NEP) aims for universal access via nationwide grid expansion, but progress in remote areas is constrained by financial limits and suspended external funding. This study evaluates the techno-economic feasibility of decentralized microgrids as an alternative to conventional grid extension under current budgetary conditions. We integrate a terrain-adjusted MV line-cost model with (i) PLEXOS capacity expansion and chronological dispatch for centralized supply and (ii) HOMER Pro optimization for PV–diesel–battery microgrids. Key indicators include LCOE, NPC, CAPEX, OPEX, reliability (ASAI/max shortage), renewable fraction, and unserved energy. Sensitivity analyses cover diesel, PV, and battery prices, as well as discount rate variations. The results show microgrids are more cost-effective in terrain-constrained regions such as Chin State, particularly when accounting for transmission and delayed generation costs, whereas grid extension remains preferable in flat, accessible regions like Nay Pyi Taw. Diesel price is the dominant cost driver across both regions, while battery cost and discount rate affect Chin State more, and PV cost is critical in Nay Pyi Taw’s solar-rich context. These findings provide evidence-based guidance for rural electrification strategies in Myanmar and other developing countries facing similar financial and infrastructural challenges.

1. Introduction

Access to reliable and affordable electricity is a key driver of economic growth, social development, and improved living standards [1,2]. In Myanmar, national electrification remains low, with large disparities between urban and rural communities. Power generation is dominated by hydropower and natural gas, with minor contributions from other renewable sources [3]. As of December 2022, the national grid’s installed capacity was ~7.1 GW, with natural gas at ~3567 MW (~50%) and hydropower ~3225 MW (~45%); utility-scale solar and coal accounted for ~192 MW and ~138 MW, respectively [4]. In 2023, actual generation was dominated by hydropower (~57.8%), with natural gas supplying ~29–30% [5].

Household electrification improved from ~39% in 2017 to ~61.6% in 2021, with population-level access reaching ~76.8% in 2023. However, urban–rural gaps persist (urban ~94.6%, rural ~68.8%) [6,7,8], highlighting challenges in supplying remote communities (

Table 1. Myanmar power-sector snapshot and electrification indicators.

Indicator	Value/Notes	Year(s)	Source
Total installed capacity (grid)	≈7122 MW	December 2022	[4]
Capacity by fuel	Natural gas 3567 MW Hydropower 3225 MW; Solar 192 MW; Coal 138 MW	December 2022	[4]
Electricity generation share	Hydro ≈ 57.8%; Natural gas ≈ 29.5%; Others ≈ 12.7%	2023	[5]
Household electrification rate	39.1% (2017) → 57.9% (2020) → 61.6% (2021)	2017–2021	[4]
Population with access (WDI)	70.4% (2020); 72.5% (2021); 73.6% (2022); 76.8% (2023)	2020–2023	[6,8]
Urban/rural access	Urban ≈ 94.6%; Rural ≈ 68.8%	2023	[7]

; Figure 1).

The National Electrification Plan (NEP), launched in 2015 with World Bank support, targets universal access by 2030, primarily through grid expansion. Progress is constrained by financial limitations, challenging terrain, and political uncertainties [1,2]. Remote regions remain underserved due to high infrastructure costs and logistical barriers.

Decentralized energy systems, particularly microgrids, have emerged as viable alternatives. Microgrids integrate localized generation—often solar, biomass, or small hydropower—with storage to provide electricity without long-distance transmission. Evidence from other developing countries shows microgrids can reduce costs, accelerate deployment, and improve resilience [9,10,11,12,13]. However, techno-economic comparisons with grid extension in Myanmar are limited, and results vary with geography, demand density, and policy assumptions.

Recent studies employ geospatial least-cost models (OnSSET, Network Planner) for spatial optimization [14,15], multi-criteria decision analysis (AHP, PROMETHEE) for balancing economic, technical, and social factors [16], and stochastic methods to account for demand and renewable uncertainties [17]. While effective, these approaches often lack detailed chronological dispatch and hybrid-system optimization. This study addresses this gap by combining a terrain-adjusted MV line-cost model with PLEXOS for grid expansion and HOMER Pro Version 3.18.4 for PV–diesel–battery microgrids, enabling a comprehensive assessment including reliability and unserved energy.

Beyond cost, electrification planning requires consideration of system reliability, renewable penetration, and capacity deployment timing [11,18,19]. This study conducts a comparative techno-economic analysis of grid extension versus microgrid deployment in two contrasting regions: Chin State (mountainous, dispersed settlements) and Nay Pyi Taw (flat, accessible terrain). Using HOMER Pro for microgrid modeling and PLEXOS for grid expansion, we evaluate economic indicators (LCOE, CAPEX, OPEX) and technical metrics (reliability, unserved energy (ENS), renewable fraction (REF)), with sensitivity tests on diesel, PV, and battery prices, and discount rate.

By capturing both techno-economic trade-offs and contextual constraints, this study provides policy-relevant insights for Myanmar’s rural electrification and lessons for other developing countries.

2. Literature Review

Myanmar’s NEP aims to achieve universal electricity access by 2030 through a combination of grid expansion and off-grid solutions [1]. However, implementation faces financial, logistical, and infrastructural challenges [2]. Electrification strategies generally fall into two categories: centralized grid expansion and decentralized microgrid deployment. Grid expansion provides economies of scale but requires high capital and maintenance costs, particularly in remote regions [3]. Microgrids, integrating localized renewable generation, offer a cost-effective alternative where grid extension is impractical [20].

Several developing countries, including Bangladesh and Kenya, have successfully deployed microgrids using solar photovoltaic (PV) systems with battery storage to improve energy access and resilience [20,21]. Microgrids can reduce reliance on fossil fuels, minimize transmission losses, and enhance energy security [22]. Nonetheless, financing, regulatory, and technical barriers remain significant challenges [9]. Studies indicate that microgrids become economically preferable when grid extension distances exceed certain thresholds—Korkovelos et al. [10] report ~130 km, while Bekele [11] highlights that avoiding long-distance infrastructure improves cost-effectiveness.

Grid expansion remains dominant in many countries but is highly sensitive to terrain and population density. High CAPEX and OPEX, along with logistical constraints, often limit its effectiveness [2,3]. IRENA’s cost–benefit framework emphasizes the complementary role of decentralized solutions where terrain and infrastructure barriers increase costs [9]. For example, the London School of Economics reported grid expansion costs of up to GBP 17,500 per kilometer in rural areas [12].

In Myanmar, NEP implementation has slowed due to funding suspensions and budget limitations, particularly in rural regions [1]. Microgrids provide a practical alternative, enabling faster deployment and reducing dependence on large-scale transmission infrastructure [2,23]. Case studies from Nepal, India, and other developing countries show that break-even distances for grid expansion vary with terrain, population density, and proximity to existing infrastructure [23,24]. Hybrid PV–diesel–battery microgrids can deliver reliable electricity at competitive costs in off-grid communities [25].

Comparative studies of grid extension and microgrids often assume homogeneous geography or treat them in isolation, limiting applicability [26,27]. Geographic and system assumptions significantly affect techno-economic outcomes [27]. While some research considers CAPEX, OPEX, and LCOE, fewer studies integrate sensitivity analyses for diesel prices, component costs, or discount rates, which can substantially influence microgrid viability [27,28,29]. Recent standardized LCOE data, such as Lazard’s 2025 report, further support microgrid competitiveness under favorable financing [29].

Global evidence demonstrates that microgrids provide a cost-effective and sustainable alternative to traditional grid expansion, particularly in geographically challenging or financially constrained regions [13,23]. Integrating microgrids within national electrification strategies—supported by policy frameworks and financial incentives—enhances rural electrification outcomes [25,30]. Studies from India, Nepal, Southeast Asia, and South Africa show that combining technical optimization with supportive policies and business models improves both economic viability and reliability [13,23,24,25,30,31].

Recent global syntheses provide quantitative evidence and methodological advances relevant to Myanmar. Hybrid PV–diesel–battery systems can achieve LCOE between 0.20 and 0.55 USD/kWh in typical South Asian conditions, often outperforming grid extension when MV line costs exceed 15,000–20,000 USD/km [32]. Terrain and transport costs substantially affect break-even distances [10,23]. Geospatial least-cost electrification models (OnSSET, Network Planner) and multi-criteria decision analysis (MCDA) capture geographic heterogeneity and policy trade-offs [14], while standardized LCOE benchmarks confirm the continued decline in PV and storage costs [29]. By integrating terrain-adjusted MV line costs, sensitivity analysis for key drivers, and reliability mapping, this study advances previous approaches in Myanmar’s context.

Unlike prior global models or single-country case studies, this research applies a consistent techno-economic framework tailored to Myanmar’s unique geography and policy environment. Two contrasting regions—mountainous Chin State and flat Nay Pyi Taw—are analyzed using HOMER Pro for microgrid optimization and PLEXOS for grid expansion. Beyond economic metrics (NPC, LCOE, CAPEX, OPEX), this study incorporates technological indicators, including system reliability, unserved energy (ENS), and renewable fraction (REF), while testing sensitivities to diesel price, battery cost, and discount rate [11,18,19,33].

This approach provides a nuanced framework for rural electrification decision making. By explicitly considering ENS and renewable fraction alongside traditional economic metrics, this study captures both cost-effectiveness and sustainability. Previous research, such as Malla et al. [23], focused on grid-based or micro-hydro systems in Nepal but lacked solar-based hybrids, sensitivity testing, and reliability considerations. This work extends prior studies by providing structured comparisons across terrains, applicable to countries with diverse geographic and infrastructural conditions.

3. Materials and Methods

3.1. Study Design and Workflow

This study conducts a comparative techno-economic analysis of two electrification approaches in Myanmar: centralized grid extension and decentralized microgrid deployment. Two contrasting regions were selected:

• Chin State—mountainous terrain with dispersed rural settlements and low accessibility.
• Nay Pyi Taw (NPT)—flat terrain with high accessibility and better existing infrastructure.

The methodological framework (Figure 1) integrates three analytical components:

• Medium-Voltage (MV) Line-Cost Modelling (Excel-based): Estimates capital needs for 11 kV distribution line extensions to unelectrified villages, adjusted for terrain conditions.
• Centralized Generation Expansion Modelling (PLEXOS): Simulates hydro-based generation to meet additional demand from MV extension. Outputs include NPC, LCOE, CAPEX, OPEX, ENS, and reliability indices.
• Microgrid Simulation (HOMER Pro): Optimizes hybrid PV–diesel–battery systems under the same demand profiles. Indicators include NPC, LCOE, CAPEX, OPEX, and REF. Reliability targets are aligned with grid extension scenarios.

All scenarios use identical demand, resource, and financial assumptions to ensure comparability. Simulations apply chronological hourly dispatch with numerical optimization. No hardware-in-the-loop tests were performed, as these are unnecessary for planning-level analysis. The overall methodological workflow is summarized in Figure 2.

3.2. Case Study Descriptions

3.2.1. Electrification Status of Study Regions

The electrification status of the two study regions, Chin State and Nay Pyi Taw, is shown in Figure 3.

Chin State (Terrain Region)

Chin State, in western Myanmar, is predominantly mountainous and sparsely populated, with terrain that significantly constrains infrastructure development. Covering over 36,000 km², it is one of the country’s most isolated states, with a population density of only 13 persons/km², and over 80% of residents live in small, remote villages scattered across rugged hills and mountains [34,35]. Historically, electrification has been minimal, with many communities relying on isolated diesel generators or lacking electricity entirely [36]. The NEP, supported by international donors including the World Bank, has promoted access through grid extension and off-grid renewable solutions, particularly solar PV for villages located more than 10 miles from the grid [37,38]. Despite these efforts, about 25% of villages remain unelectrified due to geographic and logistical barriers. Among electrified villages, most rely on decentralized off-grid systems.

Figure 3. Electrification status in each region: (a) map of Myanmar with Chin Province (sourced from researchgate.net); (b) location of Nay Pyi Taw in Myanmar map (sourced from Map of World; watermark retained); (c) village electrification status in Chin State (authors’ own illustrations based on [37]); (d) village electrification status in Nay Pyi Taw (authors’ own illustrations based on [37]).

Figure 3. Electrification status in each region: (a) map of Myanmar with Chin Province (sourced from researchgate.net); (b) location of Nay Pyi Taw in Myanmar map (sourced from Map of World; watermark retained); (c) village electrification status in Chin State (authors’ own illustrations based on [37]); (d) village electrification status in Nay Pyi Taw (authors’ own illustrations based on [37]).

Nay Pyi Taw (Flat Region)

Nay Pyi Taw, Myanmar’s administrative capital since 2005, is a flat, centrally located union territory with well-developed road access and infrastructure [39]. With a population exceeding one million and moderate urbanization, approximately 91% of villages are electrified—around 81% via the national grid and the remainder through off-grid systems [37,39,40,41]. The region’s established grid infrastructure supports efficient grid extension and reliable centralized electricity supply. Remaining unelectrified areas are concentrated in less developed districts, indicating targeted infrastructure expansion is still required [41].

3.2.2. Case Study Area

To assess the cost-effectiveness and technical feasibility of grid extension versus microgrid electrification, representative villages were selected from Chin State (mountainous) and Nay Pyi Taw (flat). Both short- and long-distance connections to the existing grid were considered to capture varied electrification challenges.

Short-Distance Connection (Within 5 km to 10 km of Existing Grid)

In Chin State, three villages near the main grid were selected using NEP Phase II contract data [42]:

• Lailui (82 households, 1.3 miles from the grid).
• Zatour (70 households, 1.6 miles from the grid).
• Dimlo (84 households, 2.9 miles from the grid).

These villages typify rural, mountainous communities with dispersed households and limited infrastructure, suitable for evaluating both grid extension and microgrid options.

In Nay Pyi Taw, three villages from Tatkone Township were selected [42]:

• Te Myint (445 households, 1.48 miles from the grid).
• Se To (237 households, 2.54 miles from the grid).
• Chin Su (332 households, 1.23 miles from the grid).

For comparability, household counts and grid distances in NPT were standardized to match the Chin State short-line cases, isolating terrain and solar-resource effects.

Long-Distance Connection (Approximately 15 km to 20 km from Existing Grid)

To represent remote locations, a long-distance scenario was defined for both regions. In Chin State, Kwaymwe Village Tract—Selbung (82 households), Tuimi (105 households), and Vaivet (60 households), ~21 miles from the grid [42]. In Nay Pyi Taw, Letha Village Tract—Letha (131 households), Zaletgyi (194 households), and Lepan (108 households), ~19 miles from the grid [42].

For consistency, a medium-voltage (MV) line length of 10 miles (16 km) was assumed to approximate realistic routing conditions. Household counts and annual energy consumption matched the short-distance scenario to isolate line-length impacts.

This paired design evaluates how terrain, solar potential, and proximity to existing infrastructure affect the feasibility of grid extension versus microgrid deployment.

3.2.3. Average Load Consumption per Household

A bottom-up approach was used to estimate average household electricity consumption based on appliance usage and typical behavioral patterns in rural Myanmar [19,43].

Table 2. Average daily electricity consumption per household in rural Myanmar [19,43].

Access Tier	Daily Energy Use	Monthly Use	Typical Usage
Tier 2— Basic Needs	0.3–0.5 kWh	9–15 kWh	Lighting, phone charging, radio
Tier 3— Moderate Use	0.6–1.2 kWh	18–36 kWh	+Fan, TV, small appliances
Tier 4— Productive Use	1.5–3.0 kWh	45–90 kWh	+Refrigerator, water pump, small business equipment

shows daily use by access tier.

Focusing on productive-use electrification, we assume 2.5 kWh/day per household (Tier 4), i.e., 912.5 kWh/year. For the Chin State short-distance case, this is reflected in the following:

• Lailui: 82 × 0.9125 MWh = 74.825 MWh/year.
• Zatour: 70 × 0.9125 MWh = 63.875 MWh/year.
• Dimlo: 84 × 0.9125 MWh = 76.65 MWh/year.

Total: 215.35 MWh/year.

Nay Pyi Taw uses the same standardized household counts to isolate terrain and solar-resource effects. Short- and long-distance scenarios use identical household numbers and average load consumption to ensure a fair comparison. The summary of village data for the case study areas is shown in

Table 3. Summary of village data for case study areas.

Region	Village	Real Households	Assumed Households	Real Grid Distance (Miles)	Assumed Grid Distance (Miles)	Real Annual Consumption (MWh)	Assumed Annual Consumption (MWh)
Chin State (Short)	Lailui	82	82	1.3	1.3	74.83	74.83
	Zatour	70	70	1.6	1.6	63.88	63.88
	Dimlo	84	84	2.9	2.9	76.65	76.65
Nay Pyi Taw (Short Distance)	Te Myint	445	82	1.48	1.3 *	406.56	74.83
	Se To	237	70	2.54	1.6 *	216.26	63.88
	Chin Su	332	84	1.23	2.9 *	302.89	76.65
Chin State (Long Distance)	Selbung	82	82	~20	10	74.83	74.83
	Tuimi	105	70			95.81	63.88
	Vaivet	60	84			54.75	76.65
Nay Pyi Taw (Long Distance)	Letha	131	82	~19	10	119.54	74.83
	Zaletgyi	194	70			177.03	63.88
	Lepan	108	84			98.55	76.65

* Assumed distances for Nay Pyi Taw short-line villages are matched to Chin State for fair comparison. Notes: (1) Real household and grid distance data are based on contract and census sources [42]. (2) Annual consumption per household is assumed at 0.9125 MWh/year. (3) For long-distance villages, the assumed MV line distance is 10 miles (~16 km) to reflect typical routing conditions.

.

3.3. Medium-Voltage (MV) Line-Cost Modelling

3.3.1. Scope and Assumptions

We model 11 kV MV line extensions to villages, including poles, conductors, transformers, foundations, transport, and installation. High-voltage transmission is excluded. For mountainous areas (Chin State), a terrain factor adjusts for construction challenges.

3.3.2. Cost Components

The CAPEX is expressed as the sum of material, installation, transport, and transformer costs:

$$CAPEX=C\mathrm{m}\mathrm{a}\mathrm{t} +C\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{t} +C\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{p} +C\mathrm{t}\mathrm{r}$$

(1)

Annual operation and maintenance (O&M) costs are assumed as a fixed share ω of CAPEX.

3.3.3. Financials

With discount rate r and lifetime n, the capital recovery factor (CRF) is:

$$CRF\left(r,n\right)={\displaystyle \frac{r{\left(1+r\right)}^n}{{\left(1+r\right)}^n-1}}$$

(2)

The net present cost (NPC) is as follows:

$$NPC=CAPEX+\sum _{t=1}^n{\displaystyle \frac{O\&M}{{\left(1+r\right)}^t}}$$

(3)

3.3.4. Levelized Cost (MV Only)

With delivered energy E (MWh/y), the levelized cost of electricity for MV extension is:

$$LCOE(\mathrm{M}\mathrm{V})={\displaystyle \frac{CAPEX . CRF\left(r,n\right)+O\&M}{\mathrm{E}}}$$

(4)

3.3.5. Terrain Adjustment

For mountainous terrain, both CAPEX and O&M are adjusted by a multiplicative factor:

$${\mathit{CAPEX}}_{\mathrm{terrain}}= (1+\tau ) {\mathit{CAPEX}}_{\mathrm{flat}}$$

(5)

$$O\&M_{\mathrm{terrain}}=(1+\tau ) O\&M_{\mathrm{flat}}$$

(6)

with τ = 0.30 adopted per rural MV construction norms.

3.3.6. Worked Example

A comparative case study was developed for Chin State (mountainous) and Nay Pyi Taw (flat terrain), under short (<10 km) and long (15–20 km) MV extension scenarios. Detailed bills of materials and unit costs are provided in Appendix A.

MV (11 kV) line-cost modelling was performed to assess grid extension scenarios in both regions. Centralized generation planning used PLEXOS; detailed MV/LV distribution planning was carried out in Excel due to PLEXOS limitations for distribution networks. High-voltage transmission lines were excluded, consistent with the NEP focus on MV/LV infrastructure and local generation.

3.3.7. Chin State (Terrain Region)

Given the challenging terrain, H-poles with spans of ~91 m (300 ft) reduce pole counts over long distances; 3-pole and 4-pole structures are added where needed (e.g., river crossings, poor soil). This increases material and installation costs. A generic single-line diagram based on NEP Phase II contract data was used for cost estimation and layout planning (Figure 4).

Nay Pyi Taw (Flat Region)

In flat terrain, standard suspension single poles with ~46 m (150 ft) spans are typical; H-poles are used selectively at turning points, and 3-/4-pole configurations only where necessary, reducing costs (Figure 5).

The MV extension cost model includes material, installation, and labor costs for 11 kV lines, transformers, and associated hardware.

• Base cost: Derived from NEP procurement data for flat terrain.
• Terrain adjustment for Chin State: A 30% CAPEX/OPEX terrain adjustment is applied for Chin State long-distance lines, consistent with rural-electrification guidelines recommending 20–50% increases in mountainous areas [32,44,45].
• Table 4. MV (11 kV) line-cost breakdown per km in flat and mountainous terrain.

  Cost Component	Flat Terrain [USD/km]	Mountainous Terrain [USD/km]
  Poles & crossarms	5200	6760
  Conductors	3800	4940
  Transformers	2500	3250
  Labor & civil works	4000	5200
  Total	15,500	20,150

  Source: Author’s estimate based on material and labor cost data from contract and [46,47].

  shows MV (11 kV) line-cost breakdown per km in flat and mountainous terrain. A detailed breakdown of MV line-cost components is provided in Appendix A (

  Table A1. Comparative MV line-cost estimation for Chin State and Nay Pyi Taw.

  Description	Chin State Short Distance (<10 km)	Chin State Long Distance (15–20 km)	Nay Pyi Taw Short Distance (<10 km)	Nay Pyi Taw Long Distance (15–20 km)	Unit	Source/Notes
  Total line length for 3 villages	5.8	10	5.8	10	miles	NEP Contract Agreements/Adjusted lengths
  Number of transformers (11/0.4 kV)	3	3	3	3	units	NEP Contract Agreements
  11 kV line material cost per mile	23.04	23.04	18.11	18.11	million MMK	[46]
  Transformer cost per unit (50 kVA)	2.14	2.14	2.14	2.14	million MMK	[47]
  Installation cost for 11 kV line	36.13	62.30	9.18	15.83	million MMK	Estimated based on NEP Phase 2 (Chin)/ Phase 1 (Nay Pyi Taw)
  Transformer installation cost	4.30	4.30	1.50	1.50	million MMK	NEP Contract Agreements
  Transportation cost	7.98	13.75	2.37	4.09	million MMK	NEP Contract Agreements
  Total CAPEX (before adjustment)	209,747	351,353	110,737	185,119	USD	Converted at exchange rate 1360 MMK/USD (reference rate at the time of contract)
  Adjusted CAPEX (+30% terrain factor)	—	456,759	—	—	USD	30% increase for Chin’s difficult terrain
  Annual O&M cost (Transformer + Line)	4405	9592	2215	3702	USD/year	Assumed 2% of CAPEX/year
  Adjusted OPEX (+30% terrain factor)		12,470				30% increase for Chin’s difficult terrain
  Transmission cost (3% of 1 MWh)	125.85	274.06	66.44	111.07	USD/MWh	Existing 66 kV (Chin) and 33 kV (Nay Pyi Taw) transmission assumption
  Loss cost (2% losses per 1 MWh)	83.90	182.70	44.29	74.05	USD/MWh	Included in electricity delivered cost
  Delivered energy	215.35	215.35	215.35	215.35	MWh/year	Based on village load assumptions
  Discount rate (r)	10%	10%	10%	10%	—	Standard financial assumption
  Capital Recovery Factor (CRF)	0.1175	0.1175	0.1175	0.1175	—	For 20 years lifetime, 10% discount
  LCOE (CAPEX + OPEX)	0.1349	0.3071	0.0712	0.1191	USD/kWh	Levelized cost of electricity (MV line only)
  NPC	247,234	562,882	130,529	218,205	USD	Net Present Cost (MV line only)

  Notes: (1) Chin State costs include a 30% CAPEX adjustment for difficult terrain in the long-distance scenario. (2) Nay Pyi Taw costs are based on NEP Phase I contract data and adjusted line lengths to ensure a fair comparison. (3) O&M costs are estimated as 2% of CAPEX annually for all scenarios. (4) Transmission and loss costs are assumed equal for both regions, reflecting use of existing 66 kV transmission infrastructure.

  ).

3.4. Reliability Assessment and Adjustments

Accurate assessment of supply reliability is critical for comparing centralized grid extension with a decentralized microgrid, particularly in remote terrain. Reliability adjustments were integrated into both HOMER Pro and PLEXOS simulations to support a fair techno-economic evaluation.

3.4.1. Data Sources and Assumptions

Where direct data for Chin State were unavailable, operational data from the similar Rakhine Region—exhibiting similar electrification rates, feeder lengths, and terrain characteristics—were used as a proxy.

3.4.2. Reliability of Grid Extension

Reliability was quantified using the System Average Interruption Duration Index (SAIDI) and converted to the Average Service Availability Index (ASAI) [40]:

$$SADI={\displaystyle \frac{Total outage duration (customer-minutes)}{Total number of customers}}$$

(7)

• Chin State (proxy 2023): SAIDI ≈ 2948 min/year (≈590 h/year); ASAI ≈ 93%
• Nay Pyi Taw (2023): SAIDI ≈ 48.8 min/year (9.76 h/year); ASAI ≈ 99.9%.

3.5. Centralized Generation Modelling with PLEXOS

Centralized grid extension was evaluated using the PLEXOS Integrated Energy Model, a power system optimization tool for long-term capacity planning, dispatch optimization, and cost analysis. The analysis considered two representative Myanmar regions: Chin State (mountainous, remote) and Nay Pyi Taw (flat, accessible). Consistent generation assumptions were applied across both regions to ensure a fair comparison of techno-economic performance.

3.5.1. Input Data and Modelling Assumptions

A hydro-based expansion candidate (0.3 MW installed capacity) represents Myanmar’s hydropower focus (

Table 5. PLEXOS input parameters for hydro-based grid extension.

Parameter	Value	Notes/Source
Technology	Hydropower	Centralized resource
Installed Capacity (max)	0.3 MW (scaled)	Expansion candidate
Capital Cost	3250 USD/kW	[48]
Fixed O&M Cost	35 USD/kW/year	[48]
Variable O&M Cost	0 USD/MWh	No fuel cost
Fuel Cost	0	Hydro (non-fuel-based)
Capacity Factor (Max)	60%	Assumed from IEA [48]
Discount Rate	10%	Applied in cost calculations
Simulation Time Horizon	15 years	Capacity expansion & dispatch optimization
Project Lifetime	20 years	For LCOE & NPC calculations

). Demand and generation data were scaled by 10× to avoid very small system artifacts; results were normalized post-simulation. The network comprises a small hydropower plant connected via Node 01/02 (Figure 6).

3.5.2. Reliability Constraints and Mapping (LOLP/ASAI)

LOLP targets were mapped to ASAI: Chin State 7% (≈93% ASAI) and Nay Pyi Taw 0.1% (≈99.9% ASAI) (Section 3.4). Under the modeled generation scenario, these were non-binding.

3.5.3. Capacity Expansion and Chronological Dispatch Setup

A 15-year horizon (start 2024) with 2% annual demand growth was used. PLEXOS optimized hydro commissioning dates, dispatch, and ENS using hourly resolution. Construction lead times and availability factors were explicitly incorporated.

PLEXOS simulations provided hourly chronological dispatch and capacity expansion optimization but did not include real-time domain simulations.

3.5.4. Key Performance Indicators (KPIs) (PLEXOS)

The following indicators were computed and later used for comparison with microgrid modeling:

• NPC—discounted system cost.
• LCOE—USD/kWh delivered.
• CAPEX and OPEX—annualized investment and operating expenditures.
• ENS—residual unmet demand.
• ENS and generation-related NPC, CAPEX, OPEX, and LCOE were obtained directly from the software outputs. For the distribution (MV line extension), CAPEX, OPEX, NPC, and LCOE were calculated externally in Excel, incorporating reliability adjustments.

3.6. Microgrid Modelling in HOMER Pro

To assess decentralized electrification, a hybrid microgrid was simulated using HOMER Pro, an optimization tool developed by the U.S. National Renewable Energy Laboratory (NREL) for techno-economic analysis of hybrid renewable energy systems. HOMER Pro models combinations of generation sources, storage, and load profiles under varying conditions [49].

A solar PV–diesel generator–battery hybrid microgrid was selected (Figure 7), based on prior studies identifying it as cost-effective for rural Chin State, which was chosen as a test case due to its low solar irradiance [50,51].

For comparison, the same microgrid configuration was simulated for three villages in Nay Pyi Taw, representing flat terrain with higher solar potential and lower terrain-related costs. This standardized approach enables a direct evaluation of microgrid performance across contrasting geographic and resource contexts [49].

3.6.1. Solar Radiation, Load Profile, and Input Data for Microgrid Modelling

A PV–diesel–battery hybrid microgrid was simulated with HOMER Pro [49]. Chin State (Falam District) has average annual GHI ≈ 4.39 kWh/m²/day, lower than central Myanmar (5.0–5.5 kWh/m²/day), while Nay Pyi Taw was modeled with higher irradiance (~5.25 kWh/m²/day) (Figure 8) [50,51]. Custom hourly load profiles were derived from 2024 data (personal communication) and standardized annual consumption (215.35 MWh) for cross-region comparability.

In Chin State, monthly peaks occur from March to May (~0.029 MW), with minimum demand of ~0.002 MW early in the year, reflecting seasonal agricultural and lighting needs (Figure 9c). Hourly demand follows a daily cycle, peaking between 10:00 AM and 8:00 PM and reaching lows from 2:00 to 4:00 AM (Figure 9a). This custom load profile in HOMER Pro enables precise simulation of rural household energy use patterns of rural households, critical for precise techno-economic analysis.

In Nay Pyi Taw, with stable seasonal demand and higher daytime usage due to agricultural processing and small businesses (Figure 9b,d), this allows for a direct comparison of microgrid performance under contrasting resource and terrain conditions.

The hybrid microgrid comprises solar PV, diesel generators, and lead-acid batteries, with input parameters summarized in

Table 6. Input data parameters for the hybrid microgrid simulation in HOMER Pro.

Component	Parameter	Value	Notes/Source
Solar PV	Capital Cost	2500 USD/kW	HOMER default
	O&M Cost	10 USD/kW/year	HOMER default
Diesel Generator	Capital Cost	500 USD/kW	HOMER database
	Fuel Cost	0.7 USD/L	Regional diesel price
Battery (Lead Acid)	Capital Cost	300 USD/kWh	HOMER database
Project Lifetime		20 years	Same as PLEXOS
Discount Rate		10%	Same as PLEXOS

.

3.6.2. Incorporating Reliability into HOMER Pro

To reflect outages, hourly load profiles were set to zero during outage hours, replicating typical distributions. Maximum annual capacity shortage in HOMER was set to 0.1% for Nay Pyi Taw (≈99.9% ASAI) and 7% for Chin State (≈93% ASAI). A full-reliability (0%) sensitivity was also tested.

3.6.3. Simulation Assumptions, Sensitivity Parameters, and Scenarios (HOMER Pro)

To assess the economic viability of microgrid deployment versus grid extension, HOMER Pro was used to simulate a base case and perform sensitivity analyses on key techno-economic parameters on system costs and competitiveness.

Base Case Assumptions:

• Diesel price: 0.70 USD/L, based on recent local retail prices [52].
• Battery cost, 300 USD/kWh, and PV capital cost, 2500 USD/kW, sourced from HOMER Pro’s technology cost library relevant to Southeast Asia [53].
• Discount rate: 10%, reflecting Myanmar’s sovereign risk and consistent with grid extension assumptions and IEA/OECD guidance for developing economies [48].
• Project lifetime: 20 years, matching the grid extension timeframe [53].
• Maximum annual capacity shortage: 7%, corresponding to 93% system availability, aligned with Chin State reliability.

Sensitivity Parameters:

Each parameter was varied independently within realistic ranges to analyze effects on NPC, LCOE, CAPEX and OPEX:

• Diesel price: 0.70–1.30 USD/L.
• Battery cost: 150–300 USD/kWh.
• PV capital cost: 1250–2500 USD/kW.
• Discount rate: 8–12%.

Scenario Classification:

Four scenarios were defined for comparative analysis:

• Base Case: Current assumptions, as above.
• Lowest Cost Case: Favorable market trends (diesel 0.70 USD/L, battery 150 USD/kWh, PV 1250 USD/kW, discount rate 8%).
• Highest Cost Case: Pessimistic assumptions (diesel 1.30 USD/L, battery 300 USD/kWh, PV 2500 USD/kW, discount rate 12%).
• Mid Cost Case: Moderate trends (diesel 1.00 USD/L, battery 225 USD/kWh, PV 1875 USD/kW, discount rate 10%).

These scenarios provide insight into conditions under which microgrid deployment is economically preferable or not compared to grid extension.

3.6.4. Key Performance Indicators (KPIs) (HOMER Pro)

The microgrid performance indicators were selected to align with those used in PLEXOS (Section 3.5.4):

• NPC, LCOE, CAPEX, OPEX: HOMER Pro outputs.
• ENS: Directly provided by HOMER Pro.
• REF: Provided by HOMER Pro, reflecting the share of energy delivered by renewables.

NPC, CAPEX, OPEX, LCOE, ENS, and REF were directly obtained from the software, with reliability adjustments applied consistently.

3.7. Indicator Acquisition Approach

Techno-economic indicators were obtained using a combination of direct model outputs and external post-processing. Specifically, ENS, REF, and dispatch schedules were directly reported by HOMER Pro and PLEXOS. Cost-related indicators—NPC, LCOE, CAPEX, and OPEX—were calculated externally using standard discounted cash-flow formulas (Equations (1)–(6)). Reliability indicators were derived by mapping the LOLP enforced in PLEXOS into ASAI, with consistent thresholds applied in HOMER Pro. This integrated approach ensures transparency in indicator generation and comparability between grid extension and microgrid scenarios. Figure 10 illustrates model setups and clarifies which outputs are software-provided versus externally computed.

4. Results and Discussion

This study compares centralized grid extension and decentralized microgrid systems in two contrasting regions of Myanmar: Chin State (mountainous terrain) and Nay Pyi Taw (flat, solar-rich). The results highlight how geographic conditions, resource availability, and economic factors influence the technical and economic viability of different electrification strategies.

4.1. MV Line Extension Results

Based on the methodology in Section 3.3, the key outcomes of MV line extension scenarios are summarized in

Table 7. The comparisons between short-distance and long-distance extension scenarios.

Scenario	CAPEX (USD)	OPEX (USD/Year)	LCOE (USD/kWh)	NPC (USD)
Chin State Short (<10 km)	209,747	4405	0.1349	247,234
Chin State Long (15–20 km)	456,759	12,470	0.3071	562,882
Nay Pyi Taw Short (<10 km)	110,737	2215	0.0712	130,529
Nay Pyi Taw Long (15–20 km)	185,119	3702	0.1191	218,205

Note: Detailed component costs are provided in Appendix A.

. The table reports CAPEX, OPEX, LCOE, and NPC for each scenario, enabling direct comparisons of mountainous (Chin State) and flat (Nay Pyi Taw) regions, as well as short- and long-distance line extensions. These results demonstrate the significant impact of terrain and line length on electrification costs. A detailed cost breakdown is provided below, including Appendix A.

4.2. PLEXOS Simulation Results

Using the framework described in Section 3.5, PLEXOS simulations evaluated electricity generation and grid operation performance for the selected regions. The model incorporated input data, demand profiles, capacity assumptions, and reliability targets from Section 3.5. Key indicators—NPC, LCOE, CAPEX, OPEX, and ENS—were extracted to assess the economic and technical performance of the centralized grid extension option (

Table 8. Summary of PLEXOS simulation results for hydro-based grid extension.

Region	CAPEX (USD)	OPEX (USD/Year)	LCOE (USD/kWh)	NPC (USD)
Chin State	97,500	3150	0.0121	98,841
Nay Pyi Taw	97,500	3150	0.0121	98,841

).

Figure 11 shows unserved energy elimination following plant commissioning, while Figure 12 illustrates CAPEX, OPEX, and LCOE trends over the simulation horizon.

• Interpretation of results:

• The model confirmed that a single 0.3 MW hydro unit suffices to meet demand in both Chin State and Nay Pyi Taw.
• The planning horizon starts in January 2024, with hydro commissioning scheduled for mid-2025. ENS is observed in 2024–mid-2026, reflecting the gap between initial demand and available supply during construction.
• ENS is fully eliminated once the hydro unit becomes operational, emphasizing the importance of timely infrastructure deployment in rural unelectrified areas
• After commissioning, the system meets projected demand through 2038 under the assumed 2% annual growth, confirming that a single hydro unit meets reliability targets in both regions.
• Differences in LOLP (7% for Chin vs. 0.1% for Nay Pyi Taw) do not affect generation outcomes, indicating that early ENS is driven by construction lead time rather than generation inadequacy.
• In Chin State, high MV line extension costs remain the dominant factor affecting overall grid extension feasibility, rather than generation costs.

This centralized generation analysis provides the cost baseline for comparing with decentralized microgrid modelling in HOMER Pro, which better captures hybrid system dynamics, storage dispatch, and localized reliability impacts.

4.3. Total Grid Extension Cost (MV Line + Generation): Chin State vs. Nay Pyi Taw

The total grid extension cost, including medium-voltage (MV) line and generation, differs markedly between Chin State and Nay Pyi Taw due to geographic and infrastructural differences (

Table 9. Total grid extension cost and LCOE comparison for Chin State and Nay Pyi Taw.

Region	Distance	CAPEX (USD)	OPEX (USD/Year)	LCOE (USD/kWh)	NPC (USD)	Notes on Cost Drivers
Chin State	5–10	307,247	7555	0.1470	346,075
	15–20	554,259	15,620	0.3192	661,723	Terrain cost adjustment (+30%)
Nay Pyi Taw	5–10	208,237	5475	0.0833	229,370
	15–20	282,619	7038	0.1311	317,046

; Figure 13).

In Chin State, short-distance (5–10 km) grid extension to remote villages requires a total investment of USD 307,247, increasing to USD 554,259 for long-distance (15–20 km) connections. These elevated costs reflect terrain-related challenges—steep slopes, limited road access, and low settlement density—which necessitated a 30% CAPEX adjustment to MV line installation. Hydro generation, though relatively low cost, is offset by the high distribution costs, pushing the long-distance LCOE to USD 0.3192/kWh.

In contrast, Nay Pyi Taw’s flat terrain and better infrastructure enable lower MV line costs and more efficient deployment. Short-distance connections require USD 208,237 in total investment, rising to USD 282,619 for long-distance extension. Corresponding LCOEs are USD 0.0833/kWh and USD 0.1311/kWh, respectively, reflecting reduced construction complexity and continued reliance on low-cost hydro generation.

The cross-region comparison highlights the dominant influence of terrain on total grid extension costs and economic competitiveness. While flat regions like Nay Pyi Taw support cost-effective centralized electrification, Chin State’s mountainous terrain significantly increases total costs, narrowing the economic gap between grid extension and decentralized microgrids.

Grid extension in Nay Pyi Taw achieves significantly lower LCOE than in Chin State due to flat terrain and better infrastructure, while Chin’s 30% terrain cost penalty greatly increases long-distance costs.

4.4. Microgrid Results (HOMER Simulations)

Following the methodology in Section 3.6, HOMER Pro simulations generated detailed technical and economic results for each scenario, summarized in

Table 10. Microgrid simulation results across sensitivity scenarios (Chin State).

Scenario	Max Annual Capacity Shortage	NPC (USD)	CAPEX (USD)	OPEX (USD/Year)	LCOE (USD/kWh)	Renewable Fraction (%)
Microgrid—Base Case	7% (93% reliability)	429,057	119,176	31,195	0.4500	29.6
Microgrid—Base Case	0% (100% reliability)	429,057	119,176	31,195	0.4500	29.6
Microgrid—Lowest Cost	7%	409,543	139,466	23,322	0.3680	63.4
Microgrid—Mid Cost	7%	472,637	144,618	33,020	0.4980	41.2
Microgrid—Highest Cost	7%	520,121	133,395	44,818	0.6280	30.4

(Chin State) and Figure 14, and

Table 11. Microgrid simulation results across sensitivity scenarios (Nay Pyi Taw).

Scenario	Max Annual Capacity Shortage	NPC (USD)	CAPEX (USD)	OPEX (USD/Year)	LCOE (USD/kWh)	Renewable Fraction (%)
Microgrid—Base Case	0.1% (99.9% reliability)	462,515	15,500	44,999	0.4400	0
Microgrid—Base Case	0% (100% reliability)	462,515	15,500	44,999	0.4400	0
Microgrid—Lowest Cost	0.1%	412,576	134,163	24,042	0.3370	64.2
Microgrid—Mid Cost	0.1%	476,428	113,776	36,507	0.4530	37.3
Microgrid—Highest Cost	0.1%	522,161	136,754	44,665	0.5720	35.4

(Nay Pyi Taw) and Figure 15. Key metrics include NPC, CAPEX, OPEX, LCOE, and REF. Simulations incorporated input data, load profiles, generation capacities (solar PV and diesel), storage specifications, and financial assumptions described in Section 3.6. The results provide insights into the economic feasibility and operational performance of hybrid microgrids in remote, mountainous, and flat regions, enabling direct comparison with centralized grid extension scenarios.

In Chin State, the diesel price dominates cost outcomes; favorable assumptions can reduce LCOE from 0.4500 to 0.3680 USD/kWh while increasing the renewable fraction from 29.6% to 63.4%.

Figure 14. Comparison of microgrid cost cases (Chin State).

Figure 14. Comparison of microgrid cost cases (Chin State).

Key Findings:

• Reliability: No significant cost difference was observed between 93% and 100% reliability, as the system met full demand under all scenarios.
• Renewable Fraction (REF): Electricity supply varies across microgrid scenarios. In the base case, REF is 29.6%, reflecting the PV–battery–diesel combination. The lowest-cost scenario achieves the highest REF (63.4%) due to reduced battery and PV costs, enabling greater renewable integration. Mid- and highest-cost scenarios yield REF of 41.2% and 30.4%, respectively, illustrating the trade-off between system cost and renewable penetration. These results demonstrate the flexibility of microgrids to increase renewable share depending on economic and technical design choices.
• Cost Scenarios: The lowest-cost scenario yields the lowest NPC (USD 409,543) and LCOE (USD 0.368/kWh) with the highest renewable fraction (63.4%). The highest-cost scenario shows substantially higher NPC and LCOE.
• Sensitivity: Diesel price has the largest impact on NPC and LCOE, with higher diesel costs reducing economic viability. Battery and PV cost reductions moderately improved outcomes. Discount rate variations showed minor impact.

For Nay Pyi Taw, which benefits from higher solar irradiance and minimal outages, results are summarized in Table 11. Although the maximum annual capacity shortage was set to 0.1%, this had a negligible effect compared to a true 0% shortage, due to the narrow difference.

In Nay Pyi Taw, PV cost reductions have a greater impact than battery cost, with favorable assumptions lowering LCOE to 0.3370 USD/kWh and raising the renewable fraction above 64%.

Figure 15. Comparison of microgrid cost cases (Nay Pyi Taw).

Figure 15. Comparison of microgrid cost cases (Nay Pyi Taw).

Key Findings:

• Renewable Fraction (REF): Results varied notably across microgrid scenarios. The base case shows 0% renewable penetration due to system design and PV constraints. The lowest-cost scenario achieves the highest renewable fraction at 64.2%, enabled by lower PV and battery costs. Mid- and highest-cost scenarios have renewable fractions of 37.3% and 35.4%, respectively. These findings indicate that renewable integration in microgrids is highly sensitive to technology costs and system design choices, enabling flexibility to increase renewable share while maintaining reliability.
• Cost Outcomes: The lowest-cost scenario achieves LCOE of USD 0.337/kWh, and the highest-cost scenario shows LCOE at USD 0.572/kWh, reflecting sensitivity to diesel price and capital costs.
• Capital–OPEX Trade-off: Capital-intensive scenarios with higher PV and battery investments reduce long-term OPEX and LCOE despite higher upfront CAPEX.

4.5. Cost Comparison and Scenario Analysis

Table 12. Grid extension scenarios and microgrid simulation results.

Region	Distance	System Type	LCOE (USD/kWh)	CAPEX (USD)	OPEX (USD/Year)	NPC (USD)	Renewable Fraction (%)
Chin (Terrain)	Short (<10 km)	Grid Extension	0.1470	307,247	7555	346,075	100 (Hydro)
Chin (Terrain)	Long (15–20 km)	Grid Extension	0.3192	554,259	15,620	661,723	100 (Hydro)
Chin (Terrain)	—	Microgrid (Base)	0.4500	119,176	31,195	429,057	29.6 (Solar + PV + Diesel)
Nay Pyi Taw	Short (<10 km)	Grid Extension	0.0833	208,237	5475	229,370	100 (Hydro)
Nay Pyi Taw	Long (15–20 km)	Grid Extension	0.1311	282,619	7038	317,046	100 (Hydro)
Nay Pyi Taw	—	Microgrid (Base)	0.4400	15,500	44,999	462,515	0 (Solar + PV + Diesel)

summarizes key economic metrics across scenarios, combining region, distance, and system type (Figure 16).

Grid extension remains lowest cost in Nay Pyi Taw under all scenarios, while in Chin State, the cost gap narrows enough for microgrids to be competitive when terrain and transmission costs are included.

Grid extension remains the lowest-cost option in flat, accessible areas like Nay Pyi Taw, with LCOEs as low as USD 0.0833/kWh and a fully renewable share of 100% from hydro generation. In contrast, difficult terrain in Chin State increases grid extension capital costs by approximately 30%, elevating the LCOE above that of microgrid options despite the latter’s higher operational costs. Microgrids, while costlier per kWh, require substantially lower upfront capital and offer faster deployment, and they incorporate partial renewable generation (e.g., 29.6% in Chin State and 0% in Nay Pyi Taw for the base case), particularly valuable in isolated or challenging locations.

Scenario-based analyses reveal that under favorable assumptions—lower diesel, battery, and PV prices, and reduced discount rates—microgrids in terrain regions can approach or surpass grid extension cost-effectiveness. Moreover, microgrids offer the flexibility to increase the renewable energy share depending on system design and cost assumptions, allowing for targeted integration of solar and storage to enhance sustainability. However, grid extension consistently outperforms microgrids in flat areas unless significant technology cost reductions occur.

4.6. Ranking of Influential Factors Affecting LCOE

Following the HOMER simulations described in Section 3.6, a sensitivity analysis was conducted to evaluate how key parameters affect the LCOE of Chin State and Nay Pyi Taw microgrids. Factors analyzed include diesel price, battery cost, discount rate and PV cost. The analysis shows the relative impact of each factor, enabling a ranking based on their influence on LCOE (

Table 13. Ranking of influential factors based on LCOE deviation.

Region	Influential Factor	LCOE Deviation (USD/kWh)	Direction of Impact
Chin State	Diesel Price (USD 0.7–1.3/L)	+0.178	Increase
	Battery Cost (USD 300–150/kWh)	–0.014	Decrease
	Discount Rate (8–12%)	±0.018	Both directions
	PV Capital Cost (USD 2500–1250/kW)	±0.007	Both directions
Nay Pyi Taw	Diesel Price (USD 0.7–1.3/L)	+0.112	Increase
	PV Capital Cost (USD 2500–1250/kW)	–0.076	Decrease
	Discount Rate (8–12%)	±0.011	Both directions
	Battery Cost (USD 300–150/kWh)	0.000	None

; Figure 17). This provides insight into which parameters are most critical for cost-effective microgrid planning.

Key Insights:

• Diesel Price: In both regions, diesel price is the most influential factor, with Chin State experiencing a higher LCOE increase (+0.178 USD/kWh) compared to Nay Pyi Taw (+0.112 USD/kWh). This underscores the vulnerability of microgrid economics to fuel price fluctuations.
• Battery Cost: While reducing battery costs from USD 300 to USD 150/kWh led to a modest decrease in LCOE in Chin State (–0.014 USD/kWh), it had no impact in Nay Pyi Taw. This suggests that battery economics become more significant when there is substantial renewable energy integration.
• Discount Rate: Changes in the discount rate (8–12%) had a moderate effect on LCOE in both regions, with Chin State showing a wider range of ±0.018 USD/kWh compared to Nay Pyi Taw’s ±0.011 USD/kWh.
• PV Capital Cost: Lowering PV costs from USD 2500 to USD 1250/kW reduced LCOE more significantly in Nay Pyi Taw (–0.076 USD/kWh) than in Chin State (–0.007 USD/kWh), highlighting the greater impact of PV cost reductions in regions with higher renewable energy penetration.

Key cost drivers impacting microgrid LCOE are ranked (

Table 14. The key cost drivers influencing microgrid economics differ by region.

Rank	Chin State (Terrain)	Nay Pyi Taw (Flat)
1	Diesel Price	Diesel Price
2	Discount Rate	PV Cost
3	Battery Cost	Discount Rate
4	PV Cost	Battery Cost

):

Diesel price volatility dominates cost sensitivity across regions, highlighting the vulnerability of hybrid microgrids to fuel costs. Financial conditions (discount rate) have a stronger effect in mountainous terrain, while PV costs are critical in flat, solar-abundant regions.

4.7. Cross-Case Synthesis

Two consistent patterns emerge: (1) terrain and MV line length are decisive; (2) sensitivity shifts absolute costs but not relative preferences—flat areas favor grid extension, whereas mountainous areas and long lines favor microgrids. These contrasts provide a structured framework for applying results beyond the two study areas, while acknowledging that broader regional validation remains a future research need.

4.8. Discussion and Implications

Topography profoundly influences grid extension feasibility. The 30% terrain adjustment for Chin State reflects real-world construction complexity and access constraints. These constraints reduce the viability of grid extension in mountainous areas. The grid extension scenario exhibits early ENS due to commissioning delays, whereas microgrids can supply power immediately and incorporate renewables (29.6% REF in the Chin base case). While Chin microgrids show a higher LCOE than long-distance grid extension (0.4500 vs. 0.3192 USD/kWh), this narrows when full transmission and generation costs are considered and if low-cost hydro is unavailable locally; substituting diesel or gas generation would further raise costs. In contrast, microgrids require lower upfront capital costs (USD 119,176 vs. USD 554,259) and avoid long-distance infrastructure. These results support a differentiated electrification approach, favoring grid extension in flat, accessible regions and microgrids where centralized generation is technically or economically impractical.

In flat Nay Pyi Taw, existing infrastructure and low-cost hydro generation make grid extension cost-effective, providing fully renewable electricity. Microgrid renewable penetration is minimal (0% in the base case), reflecting system design and resource availability. Sensitivity analyses, however, show that declining PV/storage costs or higher fuel prices could shift competitiveness in favor of microgrids in the future.

Overall, the dominant cost drivers differ by geography: financing terms and fuel dependency are critical for terrain-based microgrids, while solar module costs and resource availability dominate in flat, solar-rich areas.

4.9. Policy Recommendations

• Dual Electrification Strategy: Formalize integrated planning of grid extension and microgrids, guided by geospatial and demographic factors.
• Renewable Incentives: Promote solar-dominant microgrids through import duty exemptions and concessional financing.
• Fast-Track Microgrid Deployment: Streamline regulations and standardize technical designs to accelerate rural electrification in remote areas.
• Affordable Financing: Expand access to low-interest loans and guarantees for public and private microgrid projects.
• Aligned Reliability Standards: Implement comparable service quality benchmarks for grid and microgrids to ensure equitable electrification.
• Enhanced Planning Tools: Increase use of integrated simulation platforms (HOMER, PLEXOS) for evidence-based multi-scenario rural electrification planning.

5. Conclusions

5.1. Summary of Findings

An integrated framework—terrain-adjusted MV line-cost modeling, PLEXOS-based grid simulation, and HOMER Pro microgrid modeling—was applied to Chin State and Nay Pyi Taw. Cross-case synthesis and sensitivity analyses revealed consistent patterns: terrain and MV line length decisively influence costs as flat, accessible areas favor grid extension; mountainous, remote areas favor microgrids. Sensitivities (diesel, PV/battery costs, discount rate) shift absolute costs but not these relative preferences.

Microgrids offer faster deployment, lower upfront capital, and partial renewable integration in challenging terrain; grid extension remains cost-effective in flat areas with centralized generation. These results support a geographically differentiated electrification strategy that balances cost, reliability, and practical feasibility based on local conditions.

5.2. Limitations

This study analyzed only two representative regions, capturing extremes of terrain and accessibility; broader validation across additional regions is needed. The analysis focused on techno-economic and reliability indicators (LCOE, CAPEX, OPEX, ENS, ASAI, renewable fraction) without incorporating environmental, social, or policy factors. Household load assumptions (≈2.5 kWh/day) do not reflect full urban–rural or seasonal variation, so absolute cost estimates may require refinement in future analyses.

5.3. Future Research Directions

Building on these findings, future studies should explore the following:

• National-Scale Optimization: Extending the analysis to all Myanmar states and regions for a comprehensive lowest-cost electrification roadmap integrating both grid and microgrid options.
• Environmental and Social Indicators: Incorporating carbon emissions, job creation, and community acceptance metrics to support holistic decision making.
• Hybrid and Transitional Systems: Investigating modular microgrids capable of evolving into grid-connected mini-grids to inform dynamic long-term planning.
• Resilience and Climate Risks: Modeling climate-related hazards (flooding, landslides, extreme weather) to enhance infrastructure robustness.
• Business Models and Ownership: Examining community-based and public–private microgrid ownership models to facilitate scalable and inclusive rural electrification.