Condensing Economiser Retrofit on a Tuen Mun Food-Plant Boiler House
A food-processing plant in Tuen Mun runs two 6 t/h saturated-steam boilers on natural gas, one duty and one standby. Annual fuel use is 8.6 GWh (thermal, HHV) at an average cost of HK$0.52/kWh. Feedwater is returned at 105°C from a pressurised deaerator; blowdown is 5%. A vendor proposes a condensing flue-gas economiser for HK$1.85 million, claiming 9% fuel savings (HK$402,000/year, 4.6-year simple payback). The vendor's one-page calculation assumes flue gas can be cooled to 50°C and quotes a 210°C stack temperature measured during a 20-minute test at 60% load. The boiler house has limited space, and the existing chimney is a single-skin stainless steel stack 18 m high. The plant operates 7,800 hours per year and has a 12-month shutdown window next year. The finance director likes the payback; the maintenance team is worried about acidic condensate and plume.
1. Assess the vendor's claimed 9% saving and identify the principal technical and measurement weaknesses. Show the reasoning you would use before accepting the proposal.
Read the worked answer
First check: what sink is available? A condensing economiser only releases latent heat when the flue gas is cooled below its water dew point, roughly 55-60°C for natural gas with normal excess air. The available sink here is feedwater at 105°C. In a counterflow economiser the gas cannot be cooled below the entering water temperature, so the claimed 50°C stack temperature is thermodynamically impossible without a separate low-temperature sink. The realistic limit is about 105-110°C, which recovers sensible heat only.
Estimate the credible saving. For natural gas, a stack at 210°C with about 15% excess air carries roughly 12-15% of fuel input as sensible and latent loss. Cooling to 110°C can recover perhaps one-third to one-half of that, i.e. 3-5% of fuel input, not 9%. On 8.6 GWh that is 0.26-0.43 GWh/year, worth HK$135,000-224,000/year at HK$0.52/kWh. The vendor's HK$402,000 implies about 0.77 GWh, which would require condensing.
Measurement weaknesses. The 210°C reading was a 20-minute test at 60% load. Stack temperature, excess air and flue flow vary strongly with load; a single spot reading cannot define annual fuel use or savings. The calculation does not state gas flow, O2/CO2, ambient temperature, fuel HHV/LHV basis, or boiler efficiency before retrofit. It also does not account for the existing stack draught and plume.
Judgement. Treat the 9% claim as unproven and probably unachievable in this configuration. Require a full-load and part-load test, flue gas analysis and a heat-balance calculation before accepting any guarantee.
2. Recommend a technically robust scope and the operating constraints you would impose. Include what you would tell the finance director about payback.
Read the worked answer
Recommended scope. Install a non-condensing economiser or a feedwater preheater designed for 105°C feedwater, not a condensing unit. Use 316L stainless steel or better for the gas-side surfaces, a bypass damper for part-load and maintenance, a condensate collection and neutralisation system, and a stack termination designed for visible plume dispersal. Because space is limited, use a compact extended-surface tube bank and confirm fan or draught capacity.
Operating constraints. Do not allow the flue-gas outlet to fall below the acid dew point. For natural gas this is less severe than for oil, but sulphur and NOx can still create acidic condensate. Keep the economiser outlet above about 120°C unless a low-temperature sink is added. Interlock the bypass on low feedwater flow or high gas temperature. Review the boiler warranty and insurance position.
Alternative if condensing is required. A condensing economiser would need a separate sink below about 50°C, such as cold make-up water heating or a process hot-water loop. That may be possible in a food plant, but it adds heat exchangers, piping and control complexity; it should be a second phase after measuring the first.
What to tell the finance director. The credible saving is about 3-4% of fuel, or HK$134,000-179,000/year, not HK$402,000. At a realistic installed cost of HK$0.9-1.1 million, simple payback is 5-8 years, before maintenance and condensate treatment. The 4.6-year payback is based on an unsupported assumption. If the board needs a sub-five-year payback, this project does not meet it on current evidence.
3. Specify a performance test and measurement-and-verification approach that would protect the plant owner if the economiser is installed.
Read the worked answer
Baseline test. Before installation, conduct a boiler efficiency test at 100%, 75% and 50% load using ASME PTC 4.1 or EN 12952-15 methods. Record fuel flow, fuel temperature and pressure, calorific value, feedwater flow and temperature, steam flow, pressure and temperature, flue-gas temperature and O2/CO2, ambient temperature and barometric pressure. Take duplicate readings after steady operation for at least 30 minutes at each load.
Acceptance test. After installation, repeat the same test points with the economiser in service and bypassed. The guarantee should be expressed as fuel input per tonne of steam at specified steam conditions, corrected to a reference ambient temperature and fuel HHV, not as a single percentage. The vendor should guarantee a minimum heat recovery at each load and a maximum gas-side pressure drop.
M&V. Use a 12-month continuous trend of fuel gas, feedwater, steam and flue temperature. Normalise for load, ambient and fuel quality. Deduct any extra fan or pump power. Inspect for corrosion and fouling at each shutdown. Make final payment conditional on the acceptance test and a six-month performance check.