Energy retrofit strategies for schools in a changing climate: a cost-optimal perspective

Keywords: Schools, Retrofit, Climate Change, Costs

Introduction

The energy renovation of existing school buildings is becoming an increasingly relevant topic, motivated by two key needs: addressing ageing infrastructure and meeting climate policy goals. On the one hand, the need for renovation stems from intrinsic reasons: schools are often housed in outdated structures - over 50% of Italian school buildings are more than 50 years old - with poor energy performance and inadequate indoor comfort conditions. These issues are the result of both natural ageing and obsolete construction practices used before energy regulations were introduced. Addressing these shortcomings is crucial to provide adequate learning environments, which is a critical need considering that students spend a large part of their day within these buildings. On the other hand, the push for renovation is also driven by external reasons. The urgent need to meet European climate targets places strong pressure on the building sector since it is responsible for around 40% of EU energy consumption. As a result, existing buildings are required to undergo deep energy renovations aimed at reducing energy consumption and thus greenhouse gas emissions, while also improving their resilience to the impacts of climate change.

As energy renovation becomes a fundamental requirement, the key question is how to approach it effectively. Indeed, the renovation process also poses a dual challenge. While improving energy efficiency is essential to mitigate climate change, building energy performance is itself expected to change due to that same climate evolution. Consequently, failing to take climate evolution into account may result in retrofit strategies that are less effective - or even counterproductive - in the long term. This, in turn, exacerbates climate change, creating a loop where inadequate retrofits contribute to higher energy use and greater environmental impact.

To accelerate emissions reductions, and supported by increasingly available financial incentives, many renovation strategies adopt highly ambitious solutions, such as super-insulated envelopes and advanced HVAC systems that go far beyond current regulatory requirements. However, the suitability of such “over-performing” approaches must be carefully evaluated. Indeed, future climate scenarios may significantly alter heating and cooling demands, making some investments less efficient or even counterproductive. This becomes particularly relevant in public-sector projects, where high investment costs may discourage implementation. Moreover, the environmental cost of retrofit interventions - including embodied energy emissions - should not be overlooked in favour of operational savings alone.

All these factors raise important questions to consider when discussing retrofit strategies.

·         Can school buildings renovated to exceed current regulatory standards offer tangible benefits when projected in future climate conditions?

·         What are the most cost-effective combinations of retrofit strategies when both current and future energy performance are considered?

·         To what extent does climate change impact long-term operational costs and the overall economic viability of deep renovation measures?

The following sections present the methodology adopted to explore these questions, followed by a detailed analysis of the results based on a case study representative of Italian school buildings.

Methodology

To address the above questions, the cost-effectiveness of different retrofit proposals was evaluated through a Life Cycle Cost analysis, in compliance with EN 15459:1-2018[*]. A school located in the Apulia Region (Southern Italy) was selected as the case study, as representative of school buildings constructed between 1945 and the mid-1970s, prior to the introduction any national energy regulations. The technical standard EN 15459:1 provides an equation to compare retrofit measures against a reference strategy, based on the total cost of the intervention (including initial investment costs, the discounted value of annual operational costs, and replacement costs). In the considered case, the reference retrofit solution was designed to meet all national regulatory requirements, including both building envelope and HVAC systems. This solution was compared to alternative retrofit options, each of which involved improving one or more parameters relative to the baseline. Six categories of interventions were considered, as shown in Table 1.

The LCC analysis was conducted over a 30-year time horizon. For the first 10 years, building performance and associated operational energy costs were evaluated based on current climate conditions. For the remaining 20 years, future climate conditions were considered, modelled according to the IPCC worst-case scenario (SSP5-8.5). Given the large number of potential combinations of retrofit solutions, the LCC analysis was conducted using an optimization-based approach, integrating the Pymoo optimization library (Python) with the EnergyPlus simulation engine. EnergyPlus was adopted to simulate energy consumptions, while a Genetic Algorithm manages the simulations, guiding the process toward the most cost-effective retrofit strategy (compared to law-compliant alternatives), without the need to simulate every possible combination. Indeed, these algorithms emulate the process of natural selection by evolving a starting population of random solutions to find optimal or near-optimal results according to the simulation outputs.

To assess the variation of findings due to the climate context, the results were analysed for three different climate zones, according to the Italian legislation: climate zone C (adopting Lecce as representative city, HDD18°C = 1122, CDD18°C = 907), climatic zone D (adopting Foggia as representative city, HDD18°C = 1566, CDD18°C = 784), climatic zone E (adopting Campobasso as representative city, HDD18°C = 2408, CDD18°C = 334)

Table 1. Interventions, parameters of variation and range of variation (law-compliant values in bold).

Results

Optimisation results

The results of the LCC analyses for all the locations investigated are illustrated in Figure 1, showing the trend across 900 retrofit combinations along with the moving average.

Figure 1. LCC utility function results and moving average for climatic zones C, D and E.

The utility function adopted in the optimization problem compares the potential retrofit alternatives against the reference case, based on their overall costs. Specifically, it evaluates both the investment costs and the operational costs associated with each solution. This approach allows for assessing whether an alternative solution yields economic savings compared to the baseline. Such savings occur when the reduction in operational costs (mainly energy costs) outweighs the additional investment required by the alternative solution. According to the utility function, negative values indicate that a retrofit option outperforms the law-compliant reference option in terms of cost-effectiveness over the 30-year analysis period. Despite the climatic differences between locations, the overall trends appear fairly consistent. Indeed, the graphs reveal that only a limited number of retrofit combinations result in negative values of the utility function, thus performing better compared to the baseline.Specifically, the warmest climate zone (zone C) exhibits the greatest number of better combinations than the baseline option, accounting for 14 solutions of the optimization problem. By contrast, in the coldest climate zone (zone E), none out of the 900 retrofit alternatives showed an economic advantage over the baseline. Finally, climate zone D ranks in the middle, accounting for 5 solutions.

Looking at the optimal solution in each location, it is notable that - except for zone E - the best performing configuration meets law-compliant values for all retrofit measures (G, W, R, S, I) but improves solar control, using a lower SHGC value of 0.22 compared to the standard 0.35. However, the overall cost savings compared to the baseline option remain significantly limited: around €15,900 in climatic zone C and €6,900 in climatic zone D, over the 30-year period.

These limited savings can largely be attributed to the performance of the baseline retrofit, which complies with current Italian energy regulations. Indeed, the baseline already ensures good energy performance in the winter season, due to effective insulation, resulting in low heating needs. Hence, further enhancements to envelope insulation (as expected from interventions G, W, R, S) do not lead to significant additional savings, making it hard to amortize the higher investment cost required. Moreover, the effectiveness of these heating-oriented measures declines over time. In the second period of analysis - when future climate projections are applied - warmer winter conditions reduce heating demand even further. As a result, the return on investment for high-performance strategies tends to decline, particularly when evaluated over a long-term horizon.

Energy consumption

A more in-depth look at the variation in energy needs is provided by Figure 2, which illustrates heating and cooling energy consumption separately for the first (P1) and second (P2) analysis periods, as well as the total cumulative consumption over 30 years (green line).Data is shown for the three cities investigated and four retrofit scenarios: the law-compliant baseline, the LCC-optimal solution, the energy-optimal solution (lowest cumulative consumption), and the energy-worst solution (highest cumulative consumption).

Figure 2. Bar chart showing annual heating (orange bar) and cooling (blue bar) consumption simulated in the first period of analysis (period P1, current climate) and in the second period of analysis (period P2, future climate), for climatic zones C, D and E. The total cumulative energy consumption over the 30-year analysis period is represented by green lines.

First of all, the impact of climate change on building energy needs is evident: in all the locations, heating consumption drops in the second period of analysis (P2), while cooling consumption rises, even though during the hottest months of the year (July and August) cooling is limited to office areas, given that schools are largely unoccupied. Clearly, despite the consumption variation is fairly similar, the energy balance resulting from this shift varies by climate zone. In warmer areas (Zone C), the rise in cooling demand outweighs the reduction in heating, leading to an overall increase in total energy use. Conversely, in colder zones (D and E), heating reductions are more significant and offset the increase in cooling needs, resulting in net energy savings. This growing divergence points out the need to integrate future climate scenarios into retrofit assessments, as assumptions based only on historical weather data can lead to misleading results.

A comparison between the baseline retrofit and the best performing LCC solution reveals the advantages of reducing the Solar Heat Gain Coefficient (SHGC). Indeed, although a lower SHGC slightly increases heating demand, it results in a noticeable reduction in cooling energy use and associated costs compared to the baseline retrofit option, thus generating overall cost benefits. Clearly, the benefit appears to be greater during the second period of analysis (simulated under future climate conditions), when cooling requirements grow due to the increase in temperatures. This trade-off leads to notable operating cost savings in warmer locations. In fact, all the optimal solutions identified across climatic zones C and D feature SHGC values lower than the regulatory threshold, regardless of differences in the other retrofit measures applied. This consistent trend underscores the growing importance of prioritizing cooling-related strategies. As expected, retrofit strategies that reduce SHGC exhibit the greatest economic benefits in warmer climate zones, where cooling demand is more substantial. This is clearly reflected in the results for Lecce (zone C), which show the highest cost savings. In intermediate zones like Foggia (zone D), the reduction in cooling demand is more limited and does not fully offset the slight increase in heating costs, making the overall economic benefit less significant. In colder areas, such as Campobasso (zone E), where baseline cooling needs are minimal, the measure proves ineffective from a cost-saving perspective. Nonetheless, looking at cumulative energy use over 30 years, the LCC-optimal solutions show only marginal differences compared to the baseline or even to the least effective cases, highlighting the limited room for improvement once buildings meet regulatory standards.

Although all individual retrofit measures offer high performance, the results shown in Figure 2make it clear that their combination plays a crucial role in determining both energy use and overall costs. For example, the LCC-optimal cases differ from the baseline scenario only by a reduced SHGC value, confirming how even a single parameter can influence cost-effectiveness. Likewise, the worst-performing energy cases differ by just one or two parameters, depending on the city, despite using otherwise high-performance components. This shows that a poorly balanced mix of measures can lead to suboptimal outcomes, increasing costs without improving performance. On the other hand, the best-performing cases in terms of energy savings achieve major reductions in heating demand, but often require excessive investments, resulting in poor cost-effectiveness. These configurations have limited impact on cooling demand, meaning the energy savings are not enough to justify the extra cost.

While only a limited number of retrofit solutions appear cost-effective compared to the baseline retrofit option, broader considerations emerge when analysing results from an energy consumption perspective. For instance, Figure 3compares the LCC findings with the total cumulative energy consumption variation between retrofit proposals and baseline retrofit over the 30-year period of analysis.

Figure 3. Total cumulative energy consumption variation between retrofit proposals and baseline retrofit over the whole period of analysis vs LCC results in all the simulations performed, for climatic zones C, D and E.

Notably, across all the location analysed, the results group into four distinct clusters, each one corresponding to a different HVAC system type. This recurring pattern suggests that the HVAC system selection is the dominant factor influencing both energy performance and investment costs. The largest cluster, located in the upper left side, includes all configurations using the baseline HVAC system (I0). These solutions show the lowest total costs, but the highest energy consumption compared to other clusters. It is worth noting that all the optimal solutions from the optimisation process belong to this group. In contrast, clusters based on HVAC systems I1 and I2 yield substantially lower consumption - roughly half that of I0 - but require much higher investments. Finally, solutions based on the HVAC system I3 form the smallest cluster, with energy performance close to I1/I2 and intermediate LCC values. Nevertheless, it should be considered that these results exclude potential contributions from renewable energy systems like photovoltaic panels. Incorporating PV could offset electricity use and reshape the LCC comparison, reducing the competitiveness of gas-based systems (like I0).

Overall, the size of each cluster reflects the influence of HVAC replacement on the whole retrofit effectiveness. Indeed, a smaller cluster size suggests that, once the HVAC system has been replaced, the impact of further retrofit measures becomes more limited.

From a cost perspective, represented by the horizontal extent, each cluster spans a similar LCC range, approximately 35 €/m², which correspond to around 100,000 €. Otherwise, from the energy perspective, a significant variation in energy consumption can be found within clusters. Cluster I0 shows the highest intra-cluster variability, with energy consumption ranging up to 150 kWh/m² between solutions. The other clusters show more uniform energy performance, with intra-cluster variability decreasing from the warmest climate zone (C) to the coldest one (E), where it nearly disappears. This trend confirms that, in the considered case, the HVAC system is the dominant factor affecting energy consumption, while other retrofit measures contribute to limited energy improvements.

Despite the law-compliant retrofit already representing an optimal retrofit solution, energy consumption can still be drastically reduced by adopting other retrofit strategies. For example, in climatic zone E, switching to a high-efficiency HVAC system could reduce energy consumption by up to 75%. Similar reductions (60–70%) are observed in the other climatic zones, although such retrofit proposals fail to offer economic benefits in the medium-term, as they result in a LCC value largely above 0. By contrast, considering the cost-optimal solutions identified through the optimisation process, they are found to achieve only modest energy savings - never exceeding 8% compared to the base case - but offer a more balanced trade-off between performance and cost. However, it should be considered that the law-compliant retrofit alone ensures very low energy consumption levels, which already guarantee significantly reduced CO₂ emissions.

Conclusions

This study presents a methodology for evaluating energy retrofit strategies from a cost-benefit perspective, while accounting for future climate change impacts. The findings highlight the need of rethinking traditional design paradigms in light of ongoing climate change, particularly as building renovation will become increasingly necessary to meet the EU decarbonization targets.

Key findings

1.    The widespread assumption that exceeding regulatory energy performance standards always yields greater benefits is not supported by the analysis. In the investigated case, the law-compliant retrofit already guarantees high energy performance, thus further enhancements result in limited or negative cost-benefit outcomes. Overperforming without a clear economic or environmental return may lead to wasted resources.

2.    Even for buildings with low occupancy during the summer - such as schools - the study highlights that rising temperatures due to climate change will make cooling demands increasingly relevant. Among the six retrofit categories analysed, only those targeting a reduction in cooling loads, specifically through improved solar control, provided measurable economic benefits over the law-compliant baseline.

3.    From the energy perspective, HVAC replacements can drastically reduce energy consumption (up to 75%), although their high initial costs are rarely offset by economic savings over the life cycle.

Emerging challenges

The study reinforces the importance of integrating future climate projections into retrofit decision-making. Neglecting these changes can lead to a misestimation of future energy needs, resulting in investments in retrofit solutions that may not remain effective or resilient in the long term.

While high-performance solutions often focus on reducing operational energy, they may substantially increase embodied energy through the use of additional materials, complex systems, and associated processes (production, transport, installation, disposal). Accordingly, future retrofit strategies must adopt a life-cycle perspective, optimizing the trade-off between operational efficiency and environmental impact, while also considering the unavoidable evolution of the climatic context.

Reference

Campagna LM; Carlucci F; Fiorito F (2025). Life Cycle Cost optimization for schools energy retrofit under climate change: methodological approach and analyses in five different climates. Energy and Buildings, vol. 335, 115561. DOI: 10.1016/j.enbuild.2025.115561.