The stated normal case does not boil the coolant solely because the gas is at 152°C and the coolant saturation temperature is 156°C. However, that 4°C nominal margin is too small to absorb measurement error, coolant pressure loss, composition variation, flow maldistribution, fouling, or a compressor upset. Set the operating limit from the calculated local coolant-side wall temperature at the worst thermal load, then verify it against the local saturation temperature throughout the printed-circuit heat exchanger (PCHE).
Temperature margin and controlling limit
The number that matters is the local coolant subcooling at the wall:
Local wall subcooling = T_sat,local - T_wall,coolant
A positive result keeps the wall below the local boiling point. A result of zero marks incipient boiling; a negative result places the wall above saturation and permits vapor generation. The downstream coolant temperature of approximately 50°C describes the mixed bulk outlet, not the hottest coolant film inside a channel.
The stated 156°C boiling point must correspond to the actual 20% MEG/water mixture at its local absolute pressure. Pressure falls along a channel, so the lowest coolant pressure may also have the lowest saturation temperature. Composition can change the boiling point, and concentration measurements must use the same basis as the property calculation.
| Quantity | Stated or calculated value | Where to read or determine it |
|---|---|---|
| Gas pressure rise |
45 barg to 160 barg
|
Compressor suction and discharge pressure instruments |
| Gas temperature entering the cooler |
152°C in the stated operating case |
Fast temperature measurement at the PCHE hot inlet |
| Coolant saturation temperature |
156°C at the stated mixture condition |
Mixture property data evaluated at local absolute pressure |
| Nominal gas-to-saturation margin | 156 - 152 = 4°C |
Derived screening value; not the wall subcooling |
| Bulk coolant outlet temperature | Approximately 50°C
|
Downstream temperature instrument |
| Screening approach | Approximately 20°C
|
Conservative operating target requiring design validation |
An approximately 20°C approach can serve as an initial operating screen, but it is not a demonstrated PCHE standard or a guarantee against local boiling. Apply it to a clearly defined pair of temperatures and document that definition. A practical conservative screen is the difference between the lowest local coolant saturation temperature and the maximum credible gas inlet temperature; the final acceptance criterion still comes from the exchanger thermal and hydraulic analysis.
Symptom interpretation
The first symptom may be a temperature trend, but boiling is a pressure-and-heat-flux event. A rising gas inlet temperature increases local heat flux. Falling coolant flow raises the coolant temperature rise and lowers its heat-transfer coefficient, while falling coolant pressure reduces the saturation temperature. These effects can combine during an upset even when the mixed outlet remains near 50°C.
Look for correlated movement rather than a single alarm: increasing hot-gas temperature, decreasing coolant flow or differential pressure, declining coolant inlet pressure, unstable exchanger pressure drop, or oscillating coolant temperature. Vapor formation can create density changes and intermittent flow restriction. Those disturbances may redistribute flow into neighboring microchannels and intensify local heating.
A stable bulk outlet temperature does not clear the exchanger. A bypassed or weakly supplied channel can run hotter while well-fed channels dominate the outlet measurement. Similarly, a slow or remotely mounted gas temperature sensor can miss the short peak that imposes the highest wall temperature.
Local wall-temperature mechanism
Heat crosses the hot-gas film, the separating metal, and the coolant film. For a local one-dimensional screening calculation:
q'' = (T_hot - T_coolant) / (1/h_hot + R_wall + 1/h_coolant)
T_wall,coolant = T_coolant + q''/h_coolant
Here, q'' is local heat flux, h_hot and h_coolant are local film coefficients, and R_wall is wall thermal resistance per unit area. Use the PCHE channel geometry, actual fluid properties, flow distribution, and fouling condition to obtain these terms.
At the normal stated condition, a passive wall cannot be hotter than the local hot gas, so a true 152°C maximum hot-fluid temperature remains below a true 156°C local coolant saturation temperature. The problem begins when either value is not local or not bounding. A gas upset above 156°C does not automatically prove boiling because the coolant film may hold the wall below saturation, but it removes the simple temperature ceiling and makes the resistance calculation mandatory.
Cross-current flow also matters. The hottest wall location depends on where the hottest gas intersects the warmest or weakest coolant channel, not necessarily at either external outlet nozzle. PCHE passage tolerances, blockage, fouling, and header distribution can create local heat-flux peaks that bulk instruments cannot resolve.
Diagnostic measurements
Build a synchronized trend covering steady operation, the transition toward the 45 barg-to-160 barg duty, and the maximum credible upset duration. Use the fastest trustworthy sampling available because thermal peaks can disappear in averaged historian data.
- Confirm the hot-gas temperature at the cooler inlet and check sensor range, response, location, and calibration uncertainty.
- Record coolant inlet and outlet temperatures, absolute pressures, flow, and exchanger differential pressure on the same time base.
- Verify the MEG concentration and calculate saturation temperature at the lowest local coolant pressure, including channel and header pressure loss.
- Compare current flow and differential pressure with clean baseline data. A changed relationship can indicate restriction, bypassing, property change, or vapor formation.
- Calculate local duty from both streams where property data and measurements permit. Reconcile the two heat balances before trusting a wall-temperature model.
- Run the PCHE thermal-hydraulic model at maximum gas temperature, minimum coolant flow, minimum coolant pressure, credible fouling, and the stated mixture composition.
If field instruments cannot resolve channel conditions, use the manufacturer’s rated model or an engineering model validated against measured inlet, outlet, duty, and pressure-drop data. The calculation must report maximum coolant-side wall temperature and its location, not only terminal temperatures.
Operating and design procedure
- Define the boundary. Specify the maximum gas inlet temperature and duration, minimum coolant flow, minimum coolant inlet pressure, maximum coolant inlet temperature, composition range, and fouled condition.
-
Calculate local saturation. Determine
T_sat,localalong the coolant path from local absolute pressure and mixture composition. -
Calculate the wall profile. Model local heat flux and
T_wall,coolantacross the cross-current passage arrangement. -
Set an operating limit. Include instrument uncertainty, model uncertainty, transient lag, pressure variation, and expected fouling. The approximately
20°Capproach is a screening target until the PCHE manufacturer validates the final margin. - Protect the boundary. Use available controls to maintain coolant flow and pressure and to limit compressor loading or gas temperature before calculated wall subcooling is exhausted.
- Define the response. Alarm before the protective limit, then unload or stop the heat input if coolant flow, pressure, or temperature crosses the approved operating envelope.
Raising coolant flow can lower bulk temperature and improve the coolant film coefficient, but its benefit must be checked against higher pressure loss. Because system pressure is already at its maximum, the usable flow increase may be limited by the resulting minimum pressure inside the PCHE.
Verification and recurring pitfalls
Verify the correction during a controlled high-load test within the approved operating envelope. Acceptance requires positive calculated wall subcooling at every modeled location, stable coolant flow and differential pressure, no unexplained pressure oscillation, and agreement between measured terminal performance and the validated model.
The recurring error is comparing the 50°C bulk outlet directly with the 156°C boiling point. This hides local film temperature and local pressure. Other pitfalls include using gauge pressure in a saturation calculation that requires absolute pressure, treating a nominal MEG percentage as a measured composition, using steady-state averages for a short compressor upset, and interpreting a 20°C approach as a universal allowance.
Nucleate boiling may increase local heat transfer, but accepting it requires explicit confirmation that the PCHE, headers, coolant circuit, controls, and mixture can handle two-phase flow without instability, dryout, erosion, vibration, or flow redistribution. Accidental boiling is not an operating strategy.
Frequently asked questions
Why does a PCHE coolant boil when the outlet is only 50°C?
The outlet sensor reads mixed bulk temperature. A restricted or poorly supplied channel can have a much hotter wall, and local pressure loss can lower its saturation temperature.
Why does a 4°C margin not prove safe operation?
The 156°C - 152°C = 4°C value omits sensor uncertainty, pressure loss, composition variation, fouling, maldistribution, and transient overshoot. Calculate T_sat,local - T_wall,coolant for the limiting location.
Why does gas above 156°C not always cause coolant boiling?
The coolant-side wall can remain below saturation when the coolant is cold and its film coefficient is high. Once the hot gas exceeds the local saturation temperature, only a local resistance and heat-flux calculation can establish the wall condition.
What PCHE temperature approach should I use?
Use approximately 20°C as a conservative screening target, then replace it with a margin based on local wall temperature, local saturation temperature, uncertainties, transients, and the manufacturer’s approved operating envelope.
Stop the high-load test if coolant flow or pressure becomes unstable, the gas temperature exceeds the approved boundary, or the calculated wall subcooling is consumed. Escalate to the PCHE manufacturer’s official technical support when the model cannot bound the local wall temperature, when two-phase operation is being considered, or when measured pressure drop and thermal duty do not match the validated design.