Parametric Building Envelope Performance Dataset for Multi-Criteria Decision Making Analysis — 240 Scenarios, Cairo Hot-Arid Climate, Egypt
Description
The dataset consists of 240 parametric building envelope design scenarios for a five-storey administrative building located in Cairo, Egypt (Köppen BWh, ASHRAE climate zone 1B). The scenarios have been created by enumerating combinations of 10 types of walls, 4 types of roofs, 4 types of glazing systems, 4 shading systems, 8 orientations (0-315°, with an increment of 45°), and 7 types of window-to-wall ratios (20-80%). Seven performance measures are included for each scenario: annual cooling energy consumption (kWh/m²/yr), peak cooling load (kW), thermal discomfort hours (h/yr) based on PMV index, mean daylight illuminance (lux), U-value (W/m²K), initial construction cost (EGP/m²), and embodied carbon dioxide equivalent emissions (kgCO₂e/m²). Values of performance measures were verified against EnergyPlus v23.1 simulations using the IWEC weather file of Cairo. MCDM ranking results using AHP, ANP, TOPSIS, VIKOR,
Files
Steps to reproduce
**Steps to Reproduce** **Step 1 – Building Prototype** A five-storey administration building (≈2,000 m²) in Cairo, Egypt (30.1°N, 31.4°E; Köppen BWh; ASHRAE 1B) was used as the reference case. Internal loads followed the DOE medium office schedule: lighting 10 W/m², equipment 12 W/m², occupancy 18 m²/person, operating hours 08:00–18:00. Simulations used the **EGY_Cairo.623660_IWEC.epw** weather file with a simplified VAV all-air cooling system and no mechanical heating. **Step 2 – Parametric Design Space** Six envelope variables were investigated: wall assembly (10 types), roof assembly (4), glazing (4), shading (4), orientation (8 directions), and window-to-wall ratio (20–80% at 10% intervals). A total of **240 realistic scenarios** were generated by excluding impractical combinations. **Step 3 – Performance Evaluation** Seven performance criteria were estimated using physics-informed regression models calibrated against **EnergyPlus v23.1** simulations and validated with **HBRC/DEBE (2022)** benchmarks: * Annual cooling energy (kWh/m²·yr) * Peak cooling load (kW) * PMV discomfort hours (h/yr) * Daylight illuminance (lux) * Envelope U-value (W/m²K) * Initial construction cost (EGP/m²) * Embodied carbon (kgCO₂e/m²) **Step 4 – AHP Weighting** Criteria weights were obtained using the Analytic Hierarchy Process (AHP) based on expert pairwise comparisons. Final weights were: cooling energy 0.20, peak load 0.15, PMV 0.20, daylight 0.10, U-value 0.15, cost 0.10, and embodied carbon 0.10. The consistency ratio was **CR = 0.000**. **Step 5 – Multi-Criteria Decision Making** Six MCDM methods were implemented in Python 3.12 (NumPy, Pandas) without external MCDM libraries: * **AHP:** weighted normalized scores * **ANP:** AHP with criterion interdependence (α = 0.15) * **TOPSIS:** distance to ideal and anti-ideal solutions * **VIKOR:** compromise ranking (v = 0.5) * **PROMETHEE II:** Gaussian preference function with net outranking flow * **ELECTRE III:** concordance/discordance analysis using q = 5%, p = 20%, v = 50% The final recommendation was based on the arithmetic mean rank across all six methods. **Step 6 – Sensitivity Analysis** Robustness was evaluated using TOPSIS through (1) single-criterion dominance, assigning ≈58.8% weight to one criterion at a time, and (2) **300 Monte Carlo** simulations with ±25% random weight perturbations. **Step 7 – SHAP-Inspired Importance Analysis** Criterion importance was quantified using a leave-one-out approach: [ \phi_j(i)=CC_{all}(i)-CC_{-j}(i) ] where (CC_{-j}) is the TOPSIS score after excluding criterion *j*. Global importance was calculated as the mean absolute contribution across all 240 scenarios, producing a **240 × 7 SHAP matrix** included in the dataset.