What Stall Is and Why the Equipment Goes Cold at Part Load

Why this matters

Stall is the condition where the pressure inside a piece of steam equipment has fallen to or below the pressure its trap must discharge into, so condensate stops leaving. It is not a fault, it is an operating point, and every modulated steam device has one. The only design questions are where that point sits on the load curve and what happens to the equipment on the wrong side of it.

Getting this wrong costs a specific kind of money: repeat trap replacements that test clean and fail again, a heat exchanger that has been rebuilt twice for a problem in its return piping, and in freezing air a split coil. Getting it right usually costs one calculation done from three readings.

Stall is a point on the load curve

The trap has to push condensate against everything downstream of it: the pressure in the return main, the vertical rise from the trap outlet to that main converted at 2.31 feet of water per psi, and the friction of the return line. Call that sum the total back pressure. It does not change with load. A return main at 5 psig is at 5 psig in April, and a six foot lift is six feet in every season.

The pressure inside the equipment does change with load, because that is how steam is controlled. Steam is the one common working fluid whose temperature is set by its pressure, so a modulating valve reduces heat output by reducing pressure and letting the saturation temperature follow it down. There is no other lever available to it.

So one side of the trap collapses as the load falls and the other side does not. The load at which they meet is the stall point. Above it the trap has motive pressure and drains normally. Below it there is nothing for any trap to work with, and the trap's family, size, orifice and condition are all irrelevant to the outcome.

The three temperatures that set it

For a heat exchanger with a constant flow of the heated fluid, a constant entering fluid temperature, and condensing steam on the other side, capacity is proportional to the difference between the steam temperature and the entering fluid temperature. Under those conditions, and only under them:

stall load fraction = (saturation temperature at total back pressure - entering fluid temperature) / (saturation temperature at design supply pressure - entering fluid temperature)

Read what is and is not in that expression. The trap is not in it. The coil's surface area is not in it, and neither is its number of rows or its design leaving temperature, because those cancel: a coil with more surface needs less steam temperature at design and also loses less capacity per degree of steam temperature, and the two effects offset exactly. What is in it is a pressure pair and one fluid temperature.

The condition is load-bearing. A preheat coil whose control holds a fixed leaving air temperature while the entering air swings with the weather is a different arrangement, and it is worked through in the coil article rather than here. Where entering temperature is constant and airflow or water flow is constant, the expression above is the one to use.

The 212 F floor is what makes low-pressure systems fragile

Here is the part that surprises people who have only ever seen the arithmetic done once.

A return that is perfectly vented to atmosphere is at 0 psig, and the saturation temperature at 0 psig is 212 F. That is the floor. There is no return arrangement that puts the number below 212 F without pulling a vacuum on the return, which is a different system entirely. So the working margin available to a modulated device is the span between its supply saturation temperature and 212 F, and that span is small at low pressure:

Every psi in the return main and every foot of lift comes out of the right-hand column. On a 10 psig system there are 27 degrees to spend, and a five foot lift alone converts to 2.2 psi, which at that end of the table is worth several of them. On a 100 psig system the same five foot lift is a rounding error against 126.

That is the whole reason low-pressure heating systems generate this problem far more often than high-pressure process systems do, and it is not because low-pressure equipment is worse. It is because the pressure to temperature relationship is steep near atmospheric, so a low-pressure system is running in the narrowest part of the range it has.

What the equipment actually does below its stall point

It does not stop. This is the detail that makes stall hard to recognise, because the symptom is not a dead coil.

Condensate accumulates and the water level inside the equipment rises until the remaining dry surface is just enough to meet the load. The device then sits there, partly flooded, delivering roughly the heat that is being asked of it. The control valve is no longer setting temperature, it is setting a water level, and it is doing that through a lag measured in minutes rather than seconds.

Four consequences follow, and they are what the reader actually sees on a call.

Control quality collapses at low load and is fine at high load. Above the stall point the device holds setpoint tightly, because the valve is controlling pressure and the response is nearly immediate. Below it the delivered temperature swings, the loop hunts, and a step increase in demand is met by a device whose active surface is a fraction of its design surface, so the outlet dips well below setpoint before the level clears. A tech who arrives during peak demand finds nothing wrong. That inversion, good at design conditions and bad at mild ones, is the field fingerprint.

Discharge temperature at the trap reads low. Water sitting against a cooler surface subcools, so a temperature check at the trap can read well below saturation on a trap that is working perfectly. Anyone testing that trap on temperature alone concludes it is failed closed and replaces it.

Hammer. Each time the valve opens against a partly flooded device, steam meets a water surface at a lower temperature, collapses into it, and the water is accelerated into the void. That is the mechanism behind the bang, and it is covered in the water hammer article.

Corrosion and, in freezing air, a split. The flooded portion holds standing condensate that is carrying dissolved gases, which is a chemistry problem the feedwater articles own. In a coil with outdoor air crossing it, that standing water is the freeze.

Worked example: a convector loop and the same math at process pressure

A modulated steam convector loop heating an office space. Design supply 5 psig. Room air entering the convector at 70 F. Return main vented, so 0 psig at the main, with a 3 foot rise from the trap outlet to the main and a friction allowance of 0.5 psi.

Total back pressure: 3 feet divided by 2.31 is 1.3 psi, plus 0.5 psi, so 1.8 psig. Saturation temperature at 1.8 psig is about 218 F. Design supply 5 psig is about 227 F.

Stall fraction: (218 - 70) divided by (227 - 70), which is 148 over 157, or about 94 percent. This loop drains only above about 94 percent of its design load, which in a heating season it reaches on a handful of mornings. It is flooded for practically its whole life, and the occupant complaint is that the space overshoots and undershoots rather than that it is cold.

Now take the lift out. With the trap falling straight into a vented receiver, total back pressure is roughly 0.5 psig, near enough the 212 F floor: (212 - 70) over (227 - 70) is 142 over 157, or about 90 percent. Removing the entire lift bought four percentage points. At 5 psig design supply there is no return arrangement that gives this loop a usable operating range, because the floor alone consumes 142 of the 157 degrees it has.

Run the identical piping against a process condition to see what changed. Design supply 100 psig, about 338 F, same 1.8 psig back pressure at about 218 F, same 70 F entering fluid: (218 - 70) over (338 - 70) is 148 over 268, or about 55 percent. The same three foot lift and the same return main now leave the equipment draining down to a bit over half load. The piping did not improve. The margin did.

The failure mode of skipping this calculation is specific and common: the convector loop above gets a trap replacement program. Every trap is changed, every one of them tests clean on the day, the complaint returns within the month, and the shop that did the work is the one holding the callback. The calculation that would have prevented it needs a gauge reading, a tape measure and a steam table.

What stall is not

Four things get diagnosed as stall and are not, and each has a different cue.

An undersized trap shows a differential that is present and adequate with condensate backing up anyway. The pressures pass the gate and the equipment still floods. That is a capacity problem, and it is the case where a capacity table is the right tool.

Air binding shows as a device that will not come up to temperature on startup and is fine once running. Air is a non-condensable that occupies surface and blocks steam from reaching the tube wall; it is cleared by venting, not by pressure. The air articles own that mechanism.

A closed or plugged path shows the same flooding at every load including full load. Stall by definition releases at high load, so a device that never drains at any load has an obstruction, a shut valve or a plugged strainer, not a stall.

A control tuning problem shows hunting at all loads rather than only below a threshold. If the loop is unstable at ninety percent of load and stable at forty, that is not stall, because stall runs the other way.

How to verify the number you calculated

Take the two pressures at the same moment, under the load that produces the complaint, rather than on the convenient day. Read the pressure inside the equipment on its own gauge, read the return main gauge, and measure the lift with a tape rather than trusting a drawing. If the equipment pressure is at or below the sum, you have confirmed stall at that operating point directly, with no model in between.

Where there is no gauge on the equipment or on the return main, that absence is the first finding and it means nobody has been able to check any trap on the system. Fitting one is not a records task: it means breaking a live joint, so isolate the steam supply and the return, let the assembly cool, drain it to a point piped away from any walkway, and confirm zero on a gauge you have just seen read something else before a wrench touches the fitting. Never judge the temperature of a steam or condensate line by hand, and never trace a suspected leak by feel; the first stretch of a steam leak is invisible.

References

• Saturated steam tables from any engineering handbook, which are the authority for every pressure and saturation temperature pair in the table above
• Steam trap and pump-trap manufacturer literature, for published stall charts and for the selection of a device that supplies its own motive pressure
• 29 CFR 1910.147, the OSHA general industry standard for the control of hazardous energy, for isolating stored steam and condensate pressure before a gauge tapping is opened
• See related: What a Steam Coil Does and Why It Freezes; What Modulating Control Does to a Trap Underneath It; What Water Hammer Actually Is in a Steam Line