Rooftop PV versus Solar Thermal at a Wan Chai Hotel
A 20-storey hotel in Wan Chai has 850 m² of roof, of which 520 m² is usable. It uses 45,000 litres per day of hot water at 60°C, currently heated by gas boilers at an average marginal gas cost of HK$0.52/kWh thermal. The roof can structurally support 1.5 kN/m² and has a waterproofing membrane due for replacement. A 100 kWp PV system would cost HK$1.4 million; a solar thermal system sized for 50% of the hot-water load would cost HK$2.2 million. An adjacent tower shades the roof from 15:00 to 18:00 daily. The hotel can export PV output under a 10-year feed-in tariff of HK$4.0/kWh, or self-consume it against a daytime tariff of HK$1.35/kWh. The PV installer quotes an unshaded specific yield of 1,200 kWh/kWp/year and assumes no shading loss. The hotel wants the lowest life-cycle cost and a credible decarbonisation story.
1. Estimate the annual PV energy and revenue under the stated assumptions, correcting for shading. Compare this with the solar thermal option on a simple financial basis.
Read the worked answer
PV yield. Unshaded specific yield is 1,200 kWh/kWp/year. The 100 kWp array would produce 120,000 kWh/year before shading. The adjacent tower shades 15:00-18:00, three hours of the afternoon. In Hong Kong, those hours contain perhaps 20-25% of daily PV energy in summer and less in winter; a reasonable annual derating is 12-18%. Use 15%: 120,000 x 0.85 = 102,000 kWh/year. Also deduct soiling, inverter and availability losses of 5-8%; the installer's 1,200 kWh/kWp is usually already a net specific yield, so do not double-count. If it is a gross figure, apply 0.93: about 95,000 kWh/year. For prudence use 95,000-102,000 kWh/year.
Revenue. At the stated FIT of HK$4.0/kWh for exported output, revenue is HK$380,000-408,000/year. If the output is self-consumed instead, the value is the avoided daytime tariff of HK$1.35/kWh, only HK$128,000-138,000/year. The FIT route is therefore much more valuable, but the connection and export agreement must be confirmed. Simple payback on HK$1.4 million is 3.4-3.7 years at the FIT, or about 10 years at self-consumption value.
Solar thermal comparison. 45,000 L/day heated from about 25°C to 60°C requires 45,000 x 4.186 x 35 / 3,600 = 1,831 kWh/day, or about 668 MWh/year. A system meeting 50% of that load delivers about 334 MWh/year of heat. At a gas boiler efficiency of 80%, it avoids 334/0.8 = 418 MWh/year of gas, worth HK$217,000/year at HK$0.52/kWh. On HK$2.2 million, simple payback is about 10 years. It also faces roof-area competition and more maintenance.
Conclusion on simple financials. PV is clearly better on payback and has a lower maintenance burden. Solar thermal has a stronger direct decarbonisation link to hot water, but at these tariffs it is not the lowest-cost option.
2. Identify the non-financial and technical constraints that could change the ranking, including roof structure, waterproofing, shading, export limits and hotel operations.
Read the worked answer
Roof structure and wind. The roof can support 1.5 kN/m². A ballasted PV system may weigh 0.15-0.25 kN/m² plus ballast; a solar thermal system with filled collectors and pipework can be heavier, perhaps 0.3-0.6 kN/m². Both may be feasible, but the structural engineer must check wind uplift, especially at the parapet and corners. The hotel is in a typhoon region; fixing details and module testing matter more than the static weight.
Waterproofing. The membrane is due for replacement. This is an opportunity to coordinate penetrations and supports, but it also means the roof cannot be opened repeatedly. A PV system with non-penetrating ballast may preserve the membrane better; solar thermal needs pipe penetrations and insulation, which are more difficult to detail.
Shading. The 15:00-18:00 shading affects both technologies, but PV output loss is not linear with shaded area; string configuration and bypass diodes matter. Solar thermal loses output too, and stagnation in summer can damage panels if the load is off. The hotel's hot-water demand peaks in the morning and evening, while solar thermal peaks at midday; storage is needed, which adds plant-room space.
Export and grid limits. The FIT may have a capacity cap or require a grid connection study. If export is limited, the PV value falls toward the self-consumption case. The hotel's daytime load may absorb much of the PV, but the FIT revenue assumed full export. Confirm the scheme rules.
Operations. PV is modular and can be cleaned and monitored remotely. Solar thermal needs scaling control, pump maintenance and collector stagnation protection. Hotel engineering staff may be more comfortable with PV. The decarbonisation story should be based on measured data, not just installed capacity.
3. Recommend a phased roof strategy that best matches the hotel's stated objectives of lowest life-cycle cost and a credible decarbonisation story.
Read the worked answer
Recommendation: phase the roof. Phase 1: install a 60-80 kWp PV array on the unshaded portion of the roof, sized to the confirmed export or self-consumption arrangement, with a structural and waterproofing review. Use a ballasted, non-penetrating mounting system where possible. At 70 kWp and about 1,020 kWh/kWp net of shading, output is about 71,000 kWh/year; at HK$4.0/kWh that is HK$284,000/year. Capex is about HK$1.0 million, payback 3.5 years. This captures the best financial return and gives a visible renewable story.
Phase 2: use the remaining roof area and the replaced waterproofing to install a smaller solar thermal system, sized for 20-25% of hot-water demand, not 50%. That avoids overcommitting roof area and storage. A 20% system delivers about 134 MWh/year of heat, avoiding about 167 MWh/year of gas, worth HK$87,000/year. Capex might be HK$1.1 million, payback 12-13 years. It is weaker financially, so it should be justified by decarbonisation and resilience, or deferred until a heat-pump or waste-heat option is evaluated.
Alternative to consider. Instead of solar thermal, use the roof PV to power a heat-pump water heater with a COP of 3.0. That would deliver about 3 kWh of heat per kWh of PV, with lower maintenance than solar thermal. It may be a better use of limited roof area and should be compared before committing to solar thermal.
Decision. Do not choose 50% solar thermal now. It has the longest payback, the greatest roof conflict and the highest maintenance risk. Start with PV, measure performance, and then add hot-water decarbonisation in a second phase with actual hotel load data.