Energy Audits Reveal How Nepal’s Factories Can Slash Emissions by Up to 90 Percent

In a modest cluster of Nepalese factories, researchers have uncovered a strikingly clear lesson for the developing world: there is no single playbook for decarbonizing industry. A new study published in Results in Engineering audited six manufacturing facilities across three sectors—dairy processing, plastic pipe production, and secondary aluminum manufacturing—and found that the effectiveness of energy-saving interventions depends almost entirely on what kind of energy a factory uses in the first place. Thermally intensive plants burning diesel, wood, and waste oil responded dramatically to waste heat recovery and electrification, cutting modeled greenhouse gas emissions by 45 to 90 percent. Electrically driven plants, by contrast, gained little from those measures but thrived under solar power integration. The findings arrive at a critical moment, as emerging economies are projected to account for 87 percent of global energy demand growth through 2030, and roughly 80 percent of the world’s untapped industrial efficiency potential lies within their borders.

Nepal offers an unusually instructive setting for this kind of analysis. Although the country contributes just 0.08 percent of global greenhouse gas emissions, its industrial sector consumes 20 percent of national energy and produces 40 percent of the country’s net carbon emissions. Manufacturing is the largest contributor to both. Yet Nepal’s electricity grid is dominated by hydropower, giving it a carbon intensity of just 0.068 kilograms of CO2-equivalent per kilowatt-hour—a figure roughly 35 times lower in lifecycle terms than coal-fired electricity. That combination of an inefficient industrial base and an exceptionally clean grid creates what the researchers describe as a rare opportunity: electrifying fossil-fueled industrial heat in Nepal delivers far deeper emissions cuts than the same retrofit would in countries reliant on coal or natural gas, where studies have shown boiler electrification can sometimes increase emissions.

The research team, led by Adity Sapkota and colleagues, selected two representative facilities from each sector, deliberately choosing plants with contrasting energy profiles. The dairy facilities ran around the clock on grid electricity supplemented by diesel and biomass, processing between roughly 2,100 and 18,900 tons of milk products annually. The plastic plants produced high-density polyethylene pipes using almost exclusively electrical power, while the metal facilities cast aluminum utensils—one with an electric induction furnace, the other with fossil-fueled reverberatory and annealing furnaces burning waste oil, LPG, and diesel. Over the 2024/25 base year, the team deployed calibrated three-phase power quality analyzers, clamp meters, infrared thermometers, and thermal imaging cameras, logging electrical loads at hourly intervals and cross-checking results against twelve months of utility bills to within a 10 percent deviation.

The baseline measurements exposed substantial efficiency gaps. Nepalese dairy plants consumed 0.22 to 0.25 kilowatt-hours per kilogram of product—better than the national average of 0.3 to 0.58, but still two to four times higher than international best practice of 0.056 to 0.12. In the metal sector, the contrast was starker still: the fossil-fueled aluminum plant recorded a specific energy consumption of 9.17 kilowatt-hours per kilogram, far above the 1.65 to 5.69 range reported internationally, while its electrically driven counterpart performed at 1.18, better than comparable facilities abroad. Thermal efficiency testing revealed why. The dairy boilers operated within normal industrial ranges at 80 to 82 percent efficiency, but the oil-fired melting furnace managed only 21 percent and the LPG-fired annealing furnace a mere 17 percent, meaning most of the fuel’s energy escaped as waste heat rather than reaching the metal.

Motor systems emerged as a universal weakness. Across the 164 motors audited, average loading ranged from 52 percent in the metal facility to 64 percent in dairy—well below the 60 percent threshold at which motor efficiency and power factor begin to decline sharply. In the plastic plants, 60 percent of motors ran underloaded, and in the metal plant the figure reached 85 percent. The consequences were visible in the numbers: the extruder drive motors in the plastic facilities consumed 0.42 to 0.55 kilowatt-hours per kilogram, three to five times the benchmark range, because oversized motors churned at low load. Using the US Department of Energy’s MEASUR tool, the team calculated that upgrading motors to higher-efficiency units yields diminishing returns past a saturation point—95 percent efficiency for dairy and metal, 91 percent for plastic—beyond which replacement costs outweigh incremental savings.

The study then applied a three-tiered scenario framework to each facility: process optimization, electrification of thermal systems, and renewable energy integration. Under the first scenario, insulating steam lines and installing a desuperheater on the refrigeration system of the large dairy plant could recover 690 megawatt-hours annually, about 15 percent of total energy, and cut boiler fuel demand by 30 percent, saving roughly $100,000 per year. In the fossil-fueled aluminum plant, preheating combustion air with recuperators promised fuel savings of 43 percent, while preheating the charge material added another 25 percent—a combined 68 percent fuel reduction from relatively conventional engineering measures. In the plastic plants, where waste heat opportunities are inherently limited, insulating extruder barrel heaters and preheating polymer feedstock delivered a more modest 3 to 5 percent of facility energy.

Electrification produced the study’s most dramatic results. Replacing the diesel boiler at the large dairy plant with a 96 percent efficient electric boiler would pay for itself in six months and save $164,000 annually, while cutting facility emissions by up to 85 percent. Yet the same retrofit at the smaller dairy plant, which burns inexpensive firewood, yielded negative savings because wood and electricity cost nearly the same in Nepal—a finding the sensitivity analysis confirmed was robust even to a 20 percent rise in wood prices. In the metal sector, electrifying the inefficient melting and annealing furnaces could save nearly 60 percent of total facility energy and more than 90 percent of emissions, though the melting furnace’s financial case proved fragile, turning negative at discount rates above 8 percent. The lesson, the authors note, is that electrification’s viability hinges on the cost of the fuel being displaced.

Solar photovoltaic integration told a different story across the sectors. Hybrid systems pairing rooftop solar with lithium iron phosphate battery storage achieved renewable penetration of 15 to 90 percent depending on the facility, but battery capital and replacement costs accounted for more than half of total system expense, pushing the levelized cost of electricity to between $0.04 and $0.17 per kilowatt-hour. Only one facility, a plastic plant, recorded an LCOE below Nepal’s grid tariff of $0.068 per kilowatt-hour, and most hybrid configurations showed negative net present value with indefinite payback periods. Strip out the batteries, however, and the economics transform: PV-only systems achieved an LCOE of $0.02 per kilowatt-hour, repaid their capital within five years at every facility, and delivered positive NPV across the board. The battery remains technically necessary to replace diesel backup during grid instability, creating a gap between technical need and financial reality that the authors argue only concessional financing or capital subsidies can close.

Taken together, the results deliver a message that resists uniform technology prescription. Fossil-fuel-dependent dairy and metal plants achieved their largest energy and emissions reductions through heat recovery and electrification, while the electrically driven plastic facilities gained most from solar integration, offsetting up to 65 percent of grid dependence. The modeled emissions cuts of 50 to 90 percent from electrification vastly exceed what comparable retrofits achieve on fossil-dominated grids, underscoring that industrial decarbonization strategy must be matched to the carbon intensity of the electricity system. The authors caution that their findings rest on two facilities per sector from a single industrial cluster, with modeled rather than verified post-implementation outcomes, and they call for broader multi-facility studies. But the policy implications travel well beyond Nepal: mandatory audit programs, sector-specific benchmarks, targeted financing for capital-intensive retrofits, and grid investment to handle rising industrial demand are all prerequisites for turning audited potential into real-world emission cuts in developing economies.

Subject of Research: Cross-sectoral energy auditing and scenario-based decarbonization assessment of dairy, plastic, and secondary aluminum manufacturing facilities in Nepal

Article Title: Cross-sectoral energy audit and scenario-based decarbonization assessment of manufacturing industries in Nepal: A comparative study of dairy, plastic, and metal sectors

Article References: Sapkota, A., Paudel, S., Joshi, A., & Adhikari, B. (2026). Cross-sectoral energy audit and scenario-based decarbonization assessment of manufacturing industries in Nepal: A comparative study of dairy, plastic, and metal sectors. Results in Engineering, 32, Article 113191. https://doi.org/10.1016/j.rineng.2026.113191

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113191

Keywords: energy audit, industrial decarbonization, Nepal, electrification, solar photovoltaics, waste heat recovery, specific energy consumption, dairy processing, aluminum manufacturing, plastic extrusion, greenhouse gas emissions, energy efficiency