The motivation is grounded in policy reality. In 2025, China’s National Administration of Government Offices issued a Work Plan for Energy and Resource Conservation in Public Institutions, calling on the sector to align its energy consumption and carbon emissions with the country’s carbon peaking and carbon neutrality goals. Yet the study’s authors point out that existing evaluation standards, management mechanisms, and technical solutions are often inadequate for the demands of a rapidly decarbonizing energy system. Building stock keeps expanding, energy use patterns keep shifting, and energy management has grown more complex as a result. Previous research has tended to focus on isolated technical dimensions—energy consumption and carbon emissions—without a comprehensive, systemic evaluation approach, leaving the energy-saving potential of public institutions poorly identified and efficiency initiatives less effective than they could be.

To close that gap, the team constructed a four-dimensional evaluation system with ten core indicators. The first dimension, energy and resource consumption, tracks per capita comprehensive energy consumption, energy consumption per unit floor area, and per capita water consumption—all cost-type indicators where higher values signal worse performance. The second, advanced guidance, captures the low-carbon transition: carbon emissions per unit floor area, the utilization rate of renewable energy and green power, and the return on investment of technological retrofits. The third, operational management, measures the maturity of energy management systems and the compliance rate of energy efficiency diagnostics for equipment. The fourth introduces scenario-specific indicators tailored to institution type—campus green space coverage and semester-to-vacation energy elasticity for schools, new energy vehicle penetration for government fleets, equipment energy efficiency ratios for hospitals, and heat recovery rates for public venues.

The technical heart of the evaluation is the CRITIC method—Criteria Importance Through Intercriteria Correlation. Unlike purely subjective weighting schemes, CRITIC determines each indicator’s weight from two statistical properties: its contrast intensity, measured by standard deviation across alternatives, and its conflict with other indicators, measured by correlation coefficients. An indicator that varies strongly across retrofit options and remains relatively independent of its peers carries more information and earns a higher weight. The authors normalized the raw evaluation matrix, computed comparative strength and inter-indicator correlations, quantified each indicator’s information content, and normalized the results into final weights. The outcome was striking: technological innovation indicators dominated, with the utilization rate of renewable energy earning the highest weight at 0.258 and carbon emissions per unit floor area close behind at 0.234. In other words, for a university campus, low-carbon technology is not one consideration among many—it is the decisive one.

The framework was tested on a university in Fujian Province, a 600-acre campus serving roughly 30,000 faculty and students, whose current energy system draws about 85 percent of its supply from the municipal grid with natural gas and solar playing supplementary roles. The team modeled the campus as an integrated energy system comprising gas boilers, rooftop photovoltaics, water-source heat pumps, a combined cooling, heating, and power unit with cascade waste heat recovery, electric cooling units, and electric vehicle charging piles, all subject to energy flow balance, carbon emission, and equipment operation constraints. Seven retrofit schemes were evaluated: a comprehensive combination optimized for the highest energy efficiency index; forced scenarios mandating near-maximum deployment of CCHP, photovoltaics, or water-source heat pumps; a management-forced scenario deploying energy management systems campus-wide; a full building envelope retrofit; and a high-efficiency equipment replacement scenario.

The results were unambiguous. The comprehensive retrofit scheme scored highest at 0.907, followed by photovoltaic installation at 0.822 and equipment replacement at 0.797. Forced single-technology scenarios and envelope retrofits lagged well behind, suggesting that while any single measure may optimize certain indicators, comprehensive improvement requires coordinated deployment. Over a 20-year operational period, the comprehensive scheme reduced net present cost by 18.4 percent, annual carbon emissions by 25.2 percent, annual primary energy consumption by 28.1 percent, and annual grid electricity purchases by 34.4 percent—cutting the campus’s dependence on the public grid dramatically. Robustness checks using TOPSIS and the entropy weight method confirmed the ranking of the top schemes, validating the CRITIC-based framework despite minor variations in intermediate positions.

Evaluation alone, however, cannot explain why efficiency projects stall or succeed. Energy efficiency improvement in public institutions is inherently a strategic interaction among three parties: the government, which formulates policy and provides subsidies; integrated energy service providers, or IESPs, which implement projects and trade carbon credits; and the public institutions themselves, which decide whether to undertake retrofits. The study constructed a tripartite evolutionary game model in which each party chooses between active and passive strategies, with payoffs derived directly from the physical energy flows and carbon accounting of the integrated energy system. The public institution’s energy-saving benefit is computed from the difference in energy procurement costs before and after retrofit, while the IESP’s carbon trading revenue is defined by the total verified emission reduction multiplied by the unit carbon price—strictly coupling economic strategy to physical operations.

Stability analysis of the replicator dynamic equations, using the Jacobian matrix and Lyapunov’s first method, revealed that the game system stabilizes at only two evolutionary equilibrium points. Numerical simulations over 50 evolutionary time steps showed that the carbon trading price is the critical lever. Using case parameters, the team calculated the exact subsidy-free carbon price threshold required to trigger IESP participation at 71.29 RMB per ton. Below that threshold—at 70 RMB per ton—the government declines to enforce incentive policies, the IESP refuses to implement the project, and the educational institution pursues the retrofit autonomously because its realized benefits outweigh the modification costs. Above the threshold—at 80 RMB per ton—the IESP actively implements the project and offers incentives to the participating institution, which reciprocates by undertaking the retrofit, even though the government remains passive.

Sensitivity analysis on initial willingness and key parameters added further nuance. Regardless of the government’s initial probability, its strategy invariably converged to non-intervention, indicating that subsidies function primarily as early-stage accelerators rather than permanent supports. Higher initial willingness among all parties correlated with faster convergence to the stable cooperative state. Variations in the magnitude of government subsidies and IESP incentives exerted no significant influence on evolutionary speed during stabilization, suggesting these financial levers are guiding and auxiliary mechanisms within the framework. Most critically, the perceived cost of retrofitting dramatically affected the rate of evolution: lower perceived costs accelerated convergence, but if perceived costs exceeded expected energy-saving benefits, the system rapidly shifted toward a non-cooperative equilibrium characterized by withdrawal from the project.

The study’s practical implications are considerable. Rather than applying uniform policies, the authors recommend that policymakers use this case-based calculation methodology to determine institution-specific dynamic policy triggers: subsidies should be activated when the carbon price falls below the institution-specific mark and systematically phased out once the market price exceeds it, minimizing redundant fiscal expenditure. IESPs, for their part, should establish dynamic carbon pricing response and incentive adjustment mechanisms, and promote models such as energy management outsourcing that lower the technical barriers for public institutions undertaking retrofits. The authors acknowledge two limitations: the empirical results rest on a single university case, and the deterministic game model may not fully capture the stochastic volatility of real energy and carbon markets. Future work should validate the framework across hospitals, government agencies, and public venues, and integrate stochastic evolutionary game frameworks—potentially by introducing Gaussian white noise into dynamic pricing parameters—to enhance robustness. For now, the message is clear: measuring efficiency rigorously and modeling stakeholder behavior realistically are twin prerequisites for turning public buildings into engines of decarbonization.

Subject of Research: Energy efficiency evaluation and stakeholder strategy optimization in public institutions using CRITIC weighting and evolutionary game theory

Article Title: Evaluation and strategic optimization of public institution energy efficiency: a hybrid framework combining CRITIC weighting and tripartite evolutionary game

Article References: Chen, W., Xiong, Y., Zhang, C., Cai, Y., Jiang, Z., Xu, Z., Yang, K., Xie, S., & Zhao, Y. (2026). Evaluation and strategic optimization of public institution energy efficiency: a hybrid framework combining CRITIC weighting and tripartite evolutionary game. Frontiers in Energy Research, 14, Article 1794357. https://doi.org/10.3389/fenrg.2026.1794357

Image Credits: AI Generated

DOI: 10.3389/fenrg.2026.1794357

Keywords: public institutions, energy efficiency, CRITIC weighting, evolutionary game theory, carbon trading, integrated energy systems, building retrofits, IESPs, government subsidies, carbon inclusion, renewable energy, decarbonization

The research, led by Chen Mingyuan, Yang Youhui, Zheng Wenbin, Li Huayuan, Xuan Peizheng, and Peng Chaoyi, tackles a structural blind spot in existing electricity markets. Most day-ahead clearing frameworks were designed around pure electricity systems and do not fully represent the multi-energy coupling that defines modern integrated energy systems, or IES. These systems coordinate the conversion and delivery of electricity, heat, cooling, and natural gas, and their combined heat-and-power units, known as CHP, produce electricity and heat simultaneously from a single fuel input. When a CHP plant must run to satisfy a heat demand, it injects electricity whether or not the grid needs it, squeezing out room for wind and solar. The authors set out to build a day-ahead clearing formulation in which supply bids, coupled energy balances, carbon costs, cross-border grid transactions, and shiftable demand all appear inside one transparent mathematical program.

Technically, the model is a linear program solved over 96 quarter-hour intervals, covering a full day at 15-minute resolution. The objective function minimizes the sum of operating costs and monetized direct carbon emissions. Operating costs include accepted generation bids from wind, photovoltaic, CHP, and gas-boiler resources, net grid purchases and sales, and the cost of shifting demand between intervals. Carbon costs are computed from output-based emission factors: 0.400 kilograms of CO2 per kilowatt-hour for CHP electricity and 0.202 kilograms per kilowatt-hour for useful boiler heat, multiplied by the prevailing carbon price. The constraint system enforces an electricity balance and a heat balance in every interval, a fixed heat-to-power ratio for the CHP unit, rated-capacity and ramp-rate limits for both the CHP plant and the gas boiler, forecast-based availability ceilings on wind and solar, tie-line import and export limits, and an energy-neutral demand-shifting product in which total shifted-in energy must equal total shifted-out energy over the day.

One elegant modeling choice deserves attention. In many market-clearing formulations, preventing simultaneous buying and selling of grid power requires a binary variable, which pushes the problem into the harder mixed-integer domain. The authors show that because the buying price exceeds the selling price in every interval of their case, simultaneous purchase and sale can never be optimal: canceling the common component strictly reduces cost while preserving all physical constraints. The optimum therefore satisfies the mutual-exclusivity condition automatically, keeping the entire model affine and solvable with standard linear-programming tools. The implementation uses Python with SciPy’s linear-programming interface and the HiGHS solver, and every solution is verified through equality residuals, inequality slacks, and bound-violation checks before results are accepted.

The numerical case draws on real public data. Hourly German electricity load, wind generation, photovoltaic generation, and day-ahead prices for 15 November 2018 come from the Open Power System Data package, whose upstream sources include ENTSO-E and German transmission-system operators. The heat-demand shape comes from the When2Heat German total-heat-demand series. These profiles are scaled to a park-level system with a peak electricity demand of 2,400 kilowatts, maximum available wind power of 1,700 kilowatts, maximum photovoltaic power of 900 kilowatts, and peak heat demand of 2,300 kilowatts. The CHP unit ranges up to 2,000 kilowatts of electrical output with a base heat-to-power ratio of 1.2, the gas boiler up to 2,600 kilowatts of heat, and the reference carbon price of 505 yuan per tonne of CO2 is converted from the European Commission’s reported 2024 EU Emissions Trading System average auction price of 64.74 euros per tonne.

The headline results are striking in their precision. In the baseline scenario, with a 350-kilowatt tie-line and a 1,000 kilowatt-hour daily shiftable-energy limit, the curtailment rate is exactly 12.70 percent. Doubling the shiftable-energy limit to 2,000 kilowatt-hours per day cuts curtailment to 9.96 percent, a 2.74 percentage-point improvement, because more demand can be pushed into renewable-rich hours. But the most dramatic effect comes from spatial flexibility: raising the tie-line capacity from 350 to 800 kilowatts eliminates curtailment entirely, achieving 100 percent renewable accommodation. In that case, all 26.927 megawatt-hours of available wind and solar energy are accepted, and 11.082 megawatt-hours are exported. The comparison cleanly separates two distinct levers: demand shifting reallocates consumption in time, while tie-line capacity creates spatial transfer capability.

Perhaps the most provocative finding concerns carbon pricing. At the reference price of 505 yuan per tonne, monetizing emissions raises the objective value from 8,843.42 yuan to 15,849.03 yuan, yet the physical dispatch and the 12.70 percent curtailment rate remain completely unchanged. The heat balance and the constrained tie-line bind before the carbon-adjusted merit order can alter technology selection. Only when the carbon price reaches 1,500 yuan per tonne does the system cross a technology-substitution threshold: curtailment falls to zero and emissions decline from 13,872.50 to 13,333.46 kilograms of CO2, dropping further to 12,126.03 kilograms at 2,000 yuan per tonne. The lesson is that a carbon signal, however politically meaningful, has no operational effect unless it is large enough to overcome binding physical constraints and the relative costs of competing technologies.

The study also quantifies how the CHP coupling configuration itself shapes renewable accommodation. Sweeping the prescribed fixed heat-to-power ratio across values of 1.0, 1.2, 1.4, and 1.6 produces curtailment rates of 33.10, 12.70, 4.72, and 1.23 percent respectively. The mechanism is intuitive: for a fixed heat-demand profile, a higher ratio means less CHP electricity is needed to deliver the same thermal output, freeing electrical balancing space for wind and photovoltaic generation. The authors are careful to note that this sweep compares alternative fixed-ratio technology configurations rather than the within-unit operating flexibility of a single plant, but the implication for planners is clear: the choice of CHP coupling parameters is a first-order determinant of how much renewable energy a coupled system can absorb.

Beyond the principal scenarios, the team solved 25 combinations of tie-line capacity and daily shiftable energy to map the joint sensitivity surface, revealing strong complementarity and diminishing returns. With a 200-kilowatt tie-line, no amount of demand shifting beyond 1,000 kilowatt-hours per day lowers curtailment below 23.33 percent, because transfer capacity remains the binding bottleneck. At 350 kilowatts, shifting reduces curtailment from 16.42 to 9.96 percent but cannot eliminate it. Zero curtailment is achieved only at sufficient combinations: 650 kilowatts with at least 1,500 kilowatt-hours of daily shifting, or 800 kilowatts with at least 500 kilowatt-hours. Additional flexibility, in other words, is valuable only when the complementary constraint has headroom.

The practical message for policymakers is that carbon pricing, CHP configuration, demand response, and transmission planning should be evaluated and coordinated jointly rather than as isolated measures. Temporal and spatial flexibility should be assessed together, with demand shifting deployed before transfer capacity becomes the dominant bottleneck and grid expansion prioritized where flexible demand already exists. As China targets 3,600 gigawatts of combined wind and solar capacity by 2030 and a non-fossil share of primary energy above 25 percent, frameworks like this one, which make the physics of coupled energy systems legible to market clearing, may determine whether the next wave of renewable capacity is actually used or simply switched off.

Subject of Research: Low-carbon day-ahead joint electricity-heat market clearing for integrated energy systems with high renewable penetration

Article Title: Low-carbon day-ahead joint market clearing for integrated energy systems with high renewable energy penetration

Article References: Mingyuan, C., Youhui, Y., Wenbin, Z., Huayuan, L., Peizheng, X., & Chaoyi, P. (2026). Low-carbon day-ahead joint market clearing for integrated energy systems with high renewable energy penetration. Energy Reports, 16, Article 109773. https://doi.org/10.1016/j.egyr.2026.109773

Image Credits: AI Generated

DOI: 10.1016/j.egyr.2026.109773

Keywords: integrated energy systems, day-ahead market clearing, renewable curtailment, carbon pricing, combined heat and power, demand response, tie-line capacity, linear programming, wind and solar integration, energy transition, market design, low-carbon dispatch