This thesis investigates the thermal comfort and energy performance of Villa Elisa, a nursing home (RSA) in Laveno Mombello (Italy), combining a critical review of thermal comfort assessment methods with a detailed dynamic building energy simulation. The literature review shows that the two most widely used comfort frameworks, the PMV–PPD model and the adaptive comfort model, are poorly suited to the population that actually occupies nursing homes, since PMV was calibrated on healthy young adults and the adaptive model presupposes a capacity for autonomous behavioural adjustment that frail, bedridden, or cognitively impaired residents often lack. Building on this gap, a detailed thermal model of Villa Elisa (31 zones) was developed in TRNSYS (Type 56), using the building’s real construction stratigraphies, internal gain profiles, and comfort settings. Rather than sizing the HVAC system from a single design-day calculation, a multi-stage simulation strategy was adopted across nine scenarios: a free-floating simulation characterizing the passive thermal behaviour of the envelope; unlimited-capacity simulations at 20°C and 24°C setpoints (with and without roof insulation) quantifying the theoretical energy demand; an iterative capacity-reduction procedure converging on a design capacity of 150 kJ/(h·m²) in cooling and 110 kJ/(h·m²) in heating; an elderly-oriented scenario raising the heating setpoint to 24°C with adjusted capacities (125/150 kJ/h·m²); and a combined roof-insulation plus capacity-limited scenario. A statistical analysis (mean, variance, percentiles, and exceedance-hour counts) of the operative temperature showed that the limited-capacity design reduces the annual variance of the operative temperature in the main living zones by approximately 93% relative to the free-floating case. Adding 10 cm of roof insulation reduces annual heating demand by 11.6% (from 15,411 to 13,628 kWh in the final capacity-limited scenario, 110/150 kJ/(h·m²) at 20°C/26°C set-points; from 15,421 to 13,637 kWh with unlimited capacity) while slightly increasing cooling demand by 2.1%, yielding a net energy saving of approximately 519 euro/year in both cases. A supplementary sensitivity analysis (Section 6.7) then tested a higher, schedule-driven occupied-hours ventilation rate (0.5 ACH, against the 0.1 ACH baseline, motivated by indoor air quality rather than modelled directly through a CO₂ balance): it improves summer comfort and reduces cooling demand, but drives several zones into a substantial winter heat deficit under the installed capacity, indicating that roughly double the heating capacity (approximately 220 kJ/(h·m²)) would be required to accommodate it without further mitigation. Pairing the same increased ventilation rate with a mechanical heat recovery unit (75% sensible effectiveness, Section 6.8) largely resolves this trade-off, restoring near-baseline energy demand and comfort while retaining most of the ventilation benefit, although a residual local winter deficit remains in one zone. A first quantitative application of the PMV-PPD model to the simulated data (Section 6.9), using the resident-specific clothing and metabolism values already defined in the model, shows that the final scenario registers as persistently “cool” in winter under this index despite meeting the operative-temperature design criterion, reinforcing the thesis's methodological choice to rely on operative-temperature thresholds rather than PMV as the primary comfort criterion for this population. The same PMV-PPD assessment repeated on the elderly-oriented scenario (24°C heating set-point, 125/150 kJ/(h·m²)) shows that this winter PMV gap is largely closed by the higher set-point, with mean PMV rising from about −1.0 to about −0.27, at the substantial energy cost already quantified for that scenario. The most significant finding concerns the small ancillary bathrooms, which register a large number of overheating hours in every scenario because cooling was deliberately not installed there, a structural design choice that directly affects the frail residents who cannot perform autonomous adaptive behaviour. The thesis concludes with practical recommendations for Villa Elisa and directions for future work, including extending the PMV-PPD assessment to every zone and scenario, applying the adaptive comfort model as a further point of comparison, and extending on-site monitoring to validate the model.
Questa tesi analizza il comfort termico e le prestazioni energetiche di Villa Elisa, una residenza sanitaria assistenziale (RSA) a Laveno Mombello, combinando una rassegna critica dei metodi di valutazione del comfort termico con una simulazione energetica dinamica dettagliata dell'edificio. La rassegna della letteratura mostra come i due modelli di comfort più diffusi, il modello PMV-PPD e il modello di comfort adattivo, siano poco adatti alla popolazione che effettivamente occupa le RSA, poiché il PMV è stato calibrato su giovani adulti sani e il modello adattivo presuppone una capacità di adattamento comportamentale autonomo che i residenti fragili, allettati o con decadimento cognitivo spesso non possiedono. A partire da questo gap, è stato sviluppato in TRNSYS (Type 56) un modello termico dettagliato di Villa Elisa (31 zone), utilizzando le stratigrafie costruttive reali dell'edificio, i profili di apporti interni e le impostazioni di comfort. Anziché dimensionare l'impianto HVAC a partire da un singolo calcolo di giorno di progetto, è stata adottata una strategia di simulazione articolata in nove scenari: una simulazione free floating che caratterizza il comportamento termico passivo dell'involucro; simulazioni a capacità illimitata a 20°C e 24°C (con e senza cappotto in copertura) per quantificare la domanda energetica teorica; una procedura iterativa di riduzione della capacità che converge su una capacità di progetto di 150 kJ/(h·m²) in raffrescamento e 110 kJ/(h·m²) in riscaldamento; uno scenario orientato agli anziani con set-point di riscaldamento a 24°C e capacità adeguate (125/150 kJ/h·m²); e uno scenario combinato con cappotto e capacità limitata. Un'analisi statistica (media, varianza, percentili e conteggio delle ore di superamento soglia) della temperatura operativa ha mostrato che la configurazione a capacità limitata riduce la varianza annua della temperatura operativa nelle zone abitative principali di circa il 93% rispetto al caso free floating. L’aggiunta di 10 cm di isolante in copertura riduce la domanda annua di riscaldamento dell’11,6% (da 15.411 a 13.628 kWh nello scenario finale a capacità limitata, 110/150 kJ/(h·m²) con set-point 20°C/26°C; da 15.421 a 13.637 kWh a capacità illimitata), a fronte di un lieve aumento della domanda di raffrescamento (+2,1%), per un risparmio netto stimato di circa 519 euro/anno in entrambi i casi. Un’analisi di sensibilità supplementare (Sezione 6.7) ha poi testato un maggiore ricambio d’aria durante le ore di occupazione (0,5 ricambi/h, contro lo 0,1 di base, motivato da considerazioni di qualità dell’aria non modellate direttamente tramite un bilancio di CO₂): questo migliora il comfort estivo e riduce la domanda di raffrescamento, ma porta diverse zone in un deficit termico invernale rilevante con la capacità installata attuale, indicando che servirebbe circa il doppio della capacità di riscaldamento (circa 220 kJ/(h·m²)) per assorbirlo senza ulteriori interventi. Abbinare lo stesso maggiore ricambio d’aria a un’unità di ventilazione meccanica con recupero di calore (efficienza sensibile del 75%, Sezione 6.8) risolve in larga parte questo compromesso, riportando la domanda energetica e il comfort quasi ai livelli di base pur mantenendo gran parte del beneficio della ventilazione, sebbene resti un deficit invernale locale residuo in una zona. Una prima applicazione quantitativa del modello PMV-PPD ai dati simulati (Sezione 6.9), con i valori di abbigliamento e metabolismo dei residenti già definiti nel modello, mostra che lo scenario finale risulta persistentemente “freddo” in inverno secondo questo indice pur rispettando il criterio di progetto basato sulla temperatura operativa, confermando la scelta metodologica della tesi di basarsi su soglie di temperatura operativa piuttosto che sul PMV come criterio di comfort primario per questa popolazione. La stessa valutazione PMV-PPD ripetuta sullo scenario orientato agli anziani (set-point di riscaldamento a 24°C, 125/150 kJ/(h·m²)) mostra che questo scarto invernale del PMV è in gran parte colmato dal set-point più alto, con il PMV medio che sale da circa −1,0 a circa −0,27, a fronte del costo energetico già quantificato per quello scenario. Il risultato più significativo riguarda i piccoli bagni ancillari, che registrano un numero elevato di ore di surriscaldamento in ogni scenario poiché il raffrescamento non vi è stato deliberatamente installato, una scelta progettuale strutturale che incide direttamente sui residenti fragili, incapaci di adottare comportamenti adattivi autonomi. La tesi si conclude con raccomandazioni pratiche per Villa Elisa e indicazioni per sviluppi futuri, tra cui l'estensione della valutazione PMV-PPD a tutte le zone e gli scenari, l'applicazione del modello di comfort adattivo come ulteriore confronto, e l'estensione del monitoraggio in sito per validare il modello.
Dynamic energy simulation and thermal comfort analysis in a nursing home (RSA): the Villa Elisa case study (Laveno Mombello)
Milesi, Vittoria
2025/2026
Abstract
This thesis investigates the thermal comfort and energy performance of Villa Elisa, a nursing home (RSA) in Laveno Mombello (Italy), combining a critical review of thermal comfort assessment methods with a detailed dynamic building energy simulation. The literature review shows that the two most widely used comfort frameworks, the PMV–PPD model and the adaptive comfort model, are poorly suited to the population that actually occupies nursing homes, since PMV was calibrated on healthy young adults and the adaptive model presupposes a capacity for autonomous behavioural adjustment that frail, bedridden, or cognitively impaired residents often lack. Building on this gap, a detailed thermal model of Villa Elisa (31 zones) was developed in TRNSYS (Type 56), using the building’s real construction stratigraphies, internal gain profiles, and comfort settings. Rather than sizing the HVAC system from a single design-day calculation, a multi-stage simulation strategy was adopted across nine scenarios: a free-floating simulation characterizing the passive thermal behaviour of the envelope; unlimited-capacity simulations at 20°C and 24°C setpoints (with and without roof insulation) quantifying the theoretical energy demand; an iterative capacity-reduction procedure converging on a design capacity of 150 kJ/(h·m²) in cooling and 110 kJ/(h·m²) in heating; an elderly-oriented scenario raising the heating setpoint to 24°C with adjusted capacities (125/150 kJ/h·m²); and a combined roof-insulation plus capacity-limited scenario. A statistical analysis (mean, variance, percentiles, and exceedance-hour counts) of the operative temperature showed that the limited-capacity design reduces the annual variance of the operative temperature in the main living zones by approximately 93% relative to the free-floating case. Adding 10 cm of roof insulation reduces annual heating demand by 11.6% (from 15,411 to 13,628 kWh in the final capacity-limited scenario, 110/150 kJ/(h·m²) at 20°C/26°C set-points; from 15,421 to 13,637 kWh with unlimited capacity) while slightly increasing cooling demand by 2.1%, yielding a net energy saving of approximately 519 euro/year in both cases. A supplementary sensitivity analysis (Section 6.7) then tested a higher, schedule-driven occupied-hours ventilation rate (0.5 ACH, against the 0.1 ACH baseline, motivated by indoor air quality rather than modelled directly through a CO₂ balance): it improves summer comfort and reduces cooling demand, but drives several zones into a substantial winter heat deficit under the installed capacity, indicating that roughly double the heating capacity (approximately 220 kJ/(h·m²)) would be required to accommodate it without further mitigation. Pairing the same increased ventilation rate with a mechanical heat recovery unit (75% sensible effectiveness, Section 6.8) largely resolves this trade-off, restoring near-baseline energy demand and comfort while retaining most of the ventilation benefit, although a residual local winter deficit remains in one zone. A first quantitative application of the PMV-PPD model to the simulated data (Section 6.9), using the resident-specific clothing and metabolism values already defined in the model, shows that the final scenario registers as persistently “cool” in winter under this index despite meeting the operative-temperature design criterion, reinforcing the thesis's methodological choice to rely on operative-temperature thresholds rather than PMV as the primary comfort criterion for this population. The same PMV-PPD assessment repeated on the elderly-oriented scenario (24°C heating set-point, 125/150 kJ/(h·m²)) shows that this winter PMV gap is largely closed by the higher set-point, with mean PMV rising from about −1.0 to about −0.27, at the substantial energy cost already quantified for that scenario. The most significant finding concerns the small ancillary bathrooms, which register a large number of overheating hours in every scenario because cooling was deliberately not installed there, a structural design choice that directly affects the frail residents who cannot perform autonomous adaptive behaviour. The thesis concludes with practical recommendations for Villa Elisa and directions for future work, including extending the PMV-PPD assessment to every zone and scenario, applying the adaptive comfort model as a further point of comparison, and extending on-site monitoring to validate the model.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_10_Milesi_Tesi.pdf
accessibile in internet per tutti
Descrizione: Testo Tesi
Dimensione
4.12 MB
Formato
Adobe PDF
|
4.12 MB | Adobe PDF | Visualizza/Apri |
|
2026_10_Milesi_Executive_Summary.pdf
accessibile in internet per tutti
Descrizione: Testo Executive Summary
Dimensione
835.81 kB
Formato
Adobe PDF
|
835.81 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/266470