Condenser water reset on a steam absorption machine is not the same optimization problem as it is on a centrifugal. On a centrifugal you are trading tower fan kW against compressor lift, and the compressor responds in seconds. On an absorption machine you are trading tower fan kW against generator steam, and the machine responds on the timescale of solution transport through the heat exchanger circuit — minutes, not seconds. Get the direction of the reset wrong and you do not just lose efficiency; you push the strong solution toward the crystallization line.
Look at the trend first. Every branch below starts with a reading, not an adjustment.
Is the machine capacity-limited or just running cold?
Take three readings simultaneously: leaving chilled water temperature, steam control valve position, and entering condenser water temperature (ECWT).
If leaving CHW is at setpoint and the steam valve is modulating below 100%, the machine has margin. That is the branch where a downward condenser water reset buys you steam and where you can consider dropping tower fans to low speed. Go to the mechanism section.
If leaving CHW is at setpoint and the steam valve is pinned wide open, you are already at the machine's capacity for the current ECWT. Raising condenser water temperature will pull the machine out of setpoint immediately. On this branch the only reset available is upward-limited: hold ECWT where it is and look at tower approach instead.
If leaving CHW is below setpoint with the valve throttled hard, the machine is overproducing on cold condenser water. This is the branch where crystallization risk lives, and it is also where a poorly ramped reset causes hunting. Check the minimum allowable ECWT in the machine's operating instructions before you touch anything.
What does entering condenser water actually change inside the machine?
The cooling water loop serves two vessels in series: the absorber, then the condenser. Absorber temperature sets the concentration the lithium bromide solution must reach to hold the evaporator pressure. Condenser temperature sets the pressure the generator must boil against.
Drop ECWT and both fall. The generator now boils at a lower pressure and a lower temperature, so it needs less steam per pound of refrigerant driven off, and the required solution concentration drops. That is the whole efficiency benefit — a real reduction in steam per ton.
The failure mode is on the same axis. Solution concentration is set by the difference between generator temperature and absorber temperature. If cooling water goes cold while the steam valve is still open — a load step down, a tower fan staging up, a restart after a trip — generator temperature stays high while absorber temperature collapses. The strong solution concentrates past its solubility limit and crystallizes in the coldest part of the circuit, which is the solution heat exchanger outlet. That is why manufacturers publish a minimum ECWT and why the machine's own control has a low-temperature limiting function. Your reset routine must sit above that limit, not fight it.
Reduced condenser water flow is a different failure than reduced condenser water temperature. Cut flow far enough and the absorber can no longer condense vapor as fast as the evaporator produces it; refrigerant accumulates and the evaporator floods, capacity falls, and the machine loses control authority. Reset temperature by staging fans. Do not reset it by throttling the condenser pump below the machine's minimum flow.
How do you measure steam input without a steam flow meter?
You cannot calculate the benefit of a reset without the input side. Three options, in order of preference:
- Condensate meter on the machine's condensate return. Positive-displacement condensate meters — Cadillac and Lincoln both build them — are simple, effectively infinite turndown, and accurate. District heating operations often have spares sitting on a shelf; borrowing or renting one for a two-week test is cheaper than a permanent steam meter installation.
- Timed condensate catch. Divert condensate downstream of the trap into a drum for a measured interval at steady load and weigh it. Flash loss at the collection point introduces a small error; for comparing two ECWT setpoints at the same load, ignore it — the error is common to both readings.
- Back-calculate from the valve. Steam supply pressure, valve Cv, and position give an estimate only. Use it to sanity-check, never to prove savings.
Convert to heat input with the latent heat at supply pressure: Q_steam (Btu/h) = m_steam (lb/h) x h_fg. Then COP = (tons x 12,000) / Q_steam. Rearranged, the steam rate you should expect is lb/h per ton = 12,000 / (COP x h_fg). A single-effect machine near COP 0.7 with h_fg around 950 Btu/lb lands near 18 lb/ton-h; a double-effect at COP 1.2 on higher-pressure steam lands roughly 11-12. Take the design steam rate off the selection sheet and use these as an order-of-magnitude check, not a target.
Does the energy balance close?
An absorption machine is a three-port heat device. Heat in at the generator plus heat in at the evaporator equals heat out at the absorber and condenser, less shell losses:
Q_steam + Q_chw = Q_cw, where Q = 500 x gpm x deltaT for water in Btu/h.
Run the balance at one steady operating point before you trust any efficiency number. If the two sides disagree by more than about 10%, one of your instruments is lying and no amount of reset tuning will produce a defensible saving. Tuning does not fix instrumentation.
| Signal | Source | Symptom of a wrong value |
|---|---|---|
| Leaving CHW temp | Chiller controller / BAS well sensor | Balance closes but reported tons are wrong; reset appears to save steam that never existed |
| CHW flow | Flow meter, or pump curve plus differential pressure | Capacity error scales directly; primary/secondary bypass flow reads as load |
| Condensate flow | Condensate meter or timed catch | COP off by the same percentage; trap leaking through inflates input |
| Steam supply pressure | Gauge at the machine inlet | Wrong h_fg, wrong Q_steam; also hides a starved generator |
| ECWT | Condenser inlet well | Reset schedule sits below the machine's crystallization margin without you seeing it |
| CW flow / deltaT | Tower loop meter | Balance will not close; masks the flooded-evaporator branch |
| Ambient wet bulb | Outdoor sensor | Reset chases a target the tower cannot physically reach; fans run at high speed continuously |
Is the fan energy saved worth the steam energy spent?
Heat rejection on an absorption machine is roughly Q_reject = Q_evap x (1 + 1/COP). At COP 0.7, a 600-ton machine rejects about 1,450 tons — over twice what an electric centrifugal of the same capacity would send to the tower. The fans are a real load and worth optimizing, but the steam side is the larger number.
Two-speed fans follow the fan laws: half speed draws roughly one-eighth the power. Dropping a cell from high to low speed removes about 87% of that motor's kW. What it costs you is tower capacity — airflow falls by half and the approach widens, so ECWT rises and the generator has to work harder.
Make the comparison in dollars at the margin. The incremental steam cost is(lb/h_low - lb/h_high) x $/lb. The incremental fan saving is the measured kW difference x $/kWh. Repeat at two or three wet-bulb bands. That table is your reset schedule.
Setting the schedule and proving it worked
- Read the minimum allowable ECWT from the machine documentation. Set the reset schedule's low limit at least 2-3°F above it. Never let the schedule command below the low limit, and never disable the machine's own low-temperature limiting.
- Reset ECWT setpoint off ambient wet bulb, not dry bulb. Dry-bulb reset asks the tower for approaches it cannot reach on humid days and pins the fans on high.
- Put a deadband and minimum on/off timers on the fan stages. Two-speed motors need the coast-down delay specified in the tower documentation when switching from high to low; skipping it shocks the drive train and burns contactors.
- Interlock the reset with the chiller's run and dilution states. On shutdown, the dilution cycle must complete before condenser water goes cold and stagnant.
- Log leaving CHW temp, ECWT, steam valve position, condensate flow, and tower kW for two weeks at 1-minute resolution.
Verification is a normalized steam rate. Bin the data by load band (for example, 300-350 tons) and plot lb steam per ton-hour against ECWT before and after. If the plot shows lower steam per ton at the new ECWT and leaving CHW never left setpoint, the reset is real. If steam per ton is flat while ECWT moved, the machine was capacity-limited the whole time and you saved only fan kW.
Stop and call the machine manufacturer's service organization before you widen the reset range if you see any of the following: generator or solution temperatures climbing while capacity falls, unexplained purge activity, or leaving CHW that will not follow the valve. Those are crystallization precursors and a control test is not the right tool. Get the published minimum ECWT, minimum condenser water flow, and low-temperature limiting parameters for your serial number in writing before committing a permanent sequence to the BAS.
FAQ
Why does lowering condenser water temperature save steam on an absorption chiller?
Colder cooling water lowers absorber and condenser temperature, which reduces both the solution concentration needed to hold evaporator pressure and the pressure the generator must boil against. The generator then needs less steam per pound of refrigerant, so steam per ton drops.
Why does an absorption chiller crystallize when condenser water gets too cold?
Solution concentration tracks the difference between generator and absorber temperature. If cooling water drops sharply while the steam valve is still open, the strong solution concentrates past its solubility limit and salts out in the coldest section, typically the solution heat exchanger outlet. Keep the reset schedule above the published minimum entering condenser water temperature and ramp it slowly.
Why does reducing condenser water flow flood the evaporator?
The absorber has to take up refrigerant vapor as fast as the evaporator boils it off. Below minimum condenser water flow the absorber warms, absorption rate falls, and refrigerant accumulates in the evaporator — capacity drops and the machine loses control authority. Reset temperature by staging tower fans, not by throttling the condenser pump.