Why Some City Districts Go Energy Positive and Others Fall Short

European cities are racing to turn entire neighbourhoods into power plants, but a new study reveals that the road to so-called positive energy districts is far rockier than policy brochures suggest. Researchers led by Alberto Brunetti, Francesco Guarino, Maurizio Cellura and Sonia Longo, publishing in Energy Reports, applied a single, tightly defined accounting framework to three very different urban districts in Bucharest, Copenhagen and Parma. Their conclusion is striking: whether a district can legitimately call itself energy positive depends less on how many solar panels are bolted to rooftops and more on where analysts draw the boundaries, which energy carriers they count and how honestly they account for the embodied carbon of the renovation itself.

Positive Energy Districts, or PEDs, are championed by JPI Urban Europe as energy-efficient, flexible urban areas that produce net zero greenhouse gas emissions while generating a local or regional surplus of renewable energy each year. The concept sits at the heart of the European Union’s climate strategy, including the mission to make one hundred cities climate neutral by 2030 and the recast Energy Performance of Buildings Directive, which will require all new buildings to be Zero Emission Buildings from 2030. Yet the definition is a reference framework rather than a binding standard, and that flexibility has produced a patchwork of incompatible claims. Some studies measure positivity in delivered energy, others in primary energy or carbon dioxide equivalents, and system boundaries are frequently left vague, making cross-case comparison nearly impossible.

The Italian team tackled this ambiguity by coupling an archetype-based urban building energy modelling workflow with explicit annual energy-balance accounting and a parallel life-cycle assessment. Building stocks were assembled by fusing GIS-derived geometry with Energy Performance Certificate data, then clustered into representative archetypes, six for Bucharest, five for Copenhagen and seven for Parma. Each archetype was simulated in OpenStudio and calibrated against EPC benchmarks, with all deviations held below the ten percent acceptance threshold. The calibrated intensities were then upscaled through the TREEPED decision-support environment to produce district-level demands and electricity-flow inventories, covering imports, on-site photovoltaic generation, exports and self-consumption.

The three case studies could hardly be more different. The Bucharest district contains 443 residential buildings erected between 1960 and 1985 across 2.5 square kilometres, heated by a methane-fuelled district heating network and cooled by heat pumps. Copenhagen’s compact 0.4 square kilometre district mixes 85 residential and non-residential buildings, including prisons and gyms, served by a biomass-fuelled district heating grid. Parma’s sprawling 4.3 square kilometre district encompasses 1,247 buildings, dominated by pre-1991 housing but also including cinemas, hospitals, schools and a university campus, all previously dependent on gas boilers. Baseline heating demand ranged from a staggering 322 gigawatt-hours in Bucharest to just 32 gigawatt-hours in Copenhagen.

Each district received an integrated retrofit package combining envelope upgrades, system conversion and photovoltaic deployment, with insulation targets drawn from national regulations: Romanian Ordinul 2641/2017 in Bucharest, Danish BR18 in Copenhagen and Italy’s Decreto Requisiti Minimi in Parma. Space-heating demand fell by 24.1 percent in Bucharest, 31.2 percent in Copenhagen and 33.4 percent in Parma. But electrification pushed electricity demand up by 17.2 percent in Copenhagen and 34.2 percent in Parma, while Copenhagen’s cooling demand surged by more than 130 percent, a reminder that retrofit responses are deeply context-dependent.

The electricity balances split into three distinct regimes. Parma alone passed the electricity-only screening gate: 25 megawatts of rooftop photovoltaics generated 31.35 gigawatt-hours against a demand of 22.57 gigawatt-hours, leaving a net annual surplus of 8.78 gigawatt-hours. Copenhagen came tantalisingly close, with 9.34 megawatts of panels producing 6.60 gigawatt-hours against 8.19 gigawatt-hours of demand, but a residual deficit remained. Bucharest, despite installing roughly 41 megawatts of capacity and generating 52.82 gigawatt-hours, stayed structurally import-dependent because its 349 gigawatt-hour demand simply overwhelmed the available roof area. The lesson is blunt: photovoltaic capacity alone is a poor descriptor of feasibility, because what matters is the ratio of usable generation surface to the demand it must offset.

Even Parma’s success comes with a crucial asterisk. A sensitivity analysis varying the usable rooftop fraction showed that at 50 percent deployment the annual balance turns negative, while 75 and 100 percent configurations pass the screen. More fundamentally, the annual surplus masks a pronounced temporal mismatch: Parma remained in deficit for four months of the year, achieved hourly surplus during only 30.4 percent of the year, and saw monthly self-sufficiency swing between 21 and 79 percent. Its self-consumption ratio of just 34.6 percent means most summer generation is exported rather than used on site. Annual positivity, the authors stress, is an accounting outcome of temporal compensation, not evidence of continuous self-sufficiency.

The life-cycle assessment delivered perhaps the study’s most sobering findings. Bucharest achieved genuine net gains, cutting total life-cycle global warming potential by about 14 percent, from 264,000 to 226,874 tonnes of carbon dioxide equivalent, because operational savings dwarfed the embodied emissions of insulation, panels and end-of-life treatment. Parma told the opposite story: use-phase emissions collapsed by 79.2 percent, yet window production alone contributed 5,090 tonnes of carbon dioxide equivalent, and the net life-cycle improvement shrank to a mere 1.24 percent, a burden transfer of 78 percentage points. Copenhagen fared worst of all, with life-cycle emissions rising 35.8 percent despite a 36.6 percent operational reduction, as photovoltaic manufacturing and disposal burdens overwhelmed the savings. Across seven impact categories the picture fragmented further, with Parma’s five additional Environmental Footprint indicators all worsening even as carbon and energy demand fell.

The Copenhagen case also exposes a subtle boundary problem with wide implications. Biomass-fed district heating may qualify as renewable by upstream fuel classification, but under a geographical boundary it remains an imported carrier crossing the district perimeter, and cannot be credited as local surplus. The authors argue that renewable origin and local self-generation must be reported as distinct attributes, and that electricity-only balances should be treated as a screening gate rather than a final verdict, triggering a multi-carrier primary-energy verification whenever a thermal carrier still crosses the boundary.

From these findings the team distilled a four-stage screening framework for replication: declare the accounting rules before any calculation; screen whether feasible within-boundary photovoltaic generation can meet annual electricity demand; verify carrier scope and apply multi-carrier checks where needed; and qualify every annual result with intra-annual matching indicators and life-cycle environmental data. The framework is explicitly a screening procedure derived from just three districts, and the authors acknowledge key limitations, including calibration against standardised EPC values rather than metered consumption, archetype aggregation that suppresses building-level detail, and a single simulated reference year. Still, the message for planners is clear and urgent: a district’s path to energy positivity is written in its geometry, its demand profile and its accounting conventions long before the first solar panel is installed.

Subject of Research: Feasibility assessment of positive energy districts through archetype-based energy modelling and life-cycle assessment in three European case studies

Article Title: Scaling positive energy districts: Lessons learned from European case studies and pathways to replication

Article References: Brunetti, A., Guarino, F., Cellura, M., & Longo, S. (2026). Scaling positive energy districts: Lessons learned from European case studies and pathways to replication. Energy Reports, 16, Article 109759. https://doi.org/10.1016/j.egyr.2026.109759

Image Credits: AI Generated

DOI: 10.1016/j.egyr.2026.109759

Keywords: positive energy districts, urban building energy modelling, photovoltaics, life-cycle assessment, building retrofit, district heating, energy self-sufficiency, burden shifting, Bucharest, Copenhagen, Parma, EU energy policy