Water-Cooled Chillers
Chiller Condenser Approach Temperature
Saturated condensing temperature minus leaving condenser water temperature, with the saturated temperature read from the condenser pressure. Then the part a fixed-threshold calculator gets wrong: no verdict against a fixed number. Approach rises and falls with load, so the only comparison is your own commissioning record, scaled to the load the machine is carrying now, and by YORK's own rule for its YK where that applies.
Why There Is No Correct Approach Number
Treat the condenser as a heat exchanger with one isothermal fluid, the condensing refrigerant. With NTU = UA / (m cp), the effectiveness is 1 − exp(−NTU), the heat rejected is Q = m cp (T_sat − T_in) times that effectiveness, and the approach T_sat − T_out works out to Q / (m cp) times exp(−NTU) / (1 − exp(−NTU)). Hold the tube surface (UA) and the water flow (m cp) and everything after Q is a constant: the approach is proportional to the load. A condenser that has lost heat transfer has a larger constant, and at a lower load its approach can still be the smaller number.
An illustration of one assumed machine, not a rating, a design value or a rule: a 1,000-ton machine at 2,400 gpm of condenser water, rejecting 15,000 Btu/h per ton at full load, with a clean UA of 900,000 Btu/h per F, chosen only so the clean and fouled rows cross between half load and full load. Every Fahrenheit value below depends on those choices and nothing else on this page does; the proportionality does not depend on them.
| Load | Clean condenser | UA at 70 percent of clean |
|---|---|---|
| 100 percent | 11.19 F | 18.10 F |
| 75 percent | 8.39 F | 13.58 F |
| 50 percent | 5.60 F | 9.05 F |
| 25 percent | 2.80 F | 4.53 F |
Read the illustration's half-load row against its clean full-load row: 9.05 F on a condenser that has lost 30 percent of its heat transfer, 11.19 F on the clean one. A calculator that compares your reading to a design approach or a rule-of-thumb band tells the owner of the fouled machine everything is fine. That is why this page refuses without a record, and why Trane writes "higher than predicted" and YORK compares to "the difference recorded for a new unit" instead of printing a value for the approach.
YORK's daily inspection for its YK allows the difference to exceed the one recorded for a new unit by up to 4 F (2.2 C); that is YORK's rule for YORK's machine, unscaled by load, and this site applies it as YORK's and not as every chiller's. The same algebra applies to that rule: at part load the record scales down and a reading inside the allowance can sit well above what the record projects to at that load, so the result prints both comparisons and says when they disagree. The one number YORK prints is an allowance over your own record, and it is applied here as YORK's.
What the Scaled Comparison Assumes
The approach is the saturated condensing temperature minus the leaving condenser water temperature, both from the same machine at the same moment; the saturated temperature comes from the condenser pressure (or is the controller's own reading of it).
The baseline is the approach recorded on this machine when it was new or its tubes were clean, and the load it was recorded at is known; the comparison is to that record, never to a number from a manual or a rule of thumb. YORK's 4 F allowance is over that record, for its YK, and is applied here as YORK's.
Condenser water flow is the same as at the baseline. Lower flow raises the approach at the same load with no fouling, so a variable-speed condenser pump or a throttled valve breaks the scaling.
Non-condensables have been ruled out. Air in the condenser raises the measured condensing pressure, so the saturated temperature read from it, and the approach with it, read high with clean tubes; on a low-pressure machine the purge run time is the check.
The load figure tracks the heat the condenser rejects. Percent load from the controller is the nearest reading to it; percent RLA is a motor current, not a heat rejection. A load entered higher than the true one raises the scaled record and can hide a rise, so enter the lower figure when two disagree.
The scaling holds the tube surface and both heat transfer coefficients constant across the load change. A real condenser's refrigerant-side coefficient rises as heat flux falls, so at loads below the one the record was taken at the scaled record errs high and a small rise can sit inside it; read the machine against a record taken at the load it is at whenever the log has one.
The two things that break the comparison from outside, air and low flow, push the approach up. A reading above the scaled record is therefore a reason to check the purge and the water flow before the tubes are cleaned; a reading at or below it, at the same flow, is not something either confounder can manufacture. The two assumptions inside the scaling, the load figure and the constant coefficients, both err the other way at part load, which is why the record taken nearest the present load is the one to read against. And one reading is not a trend: YORK names the signal as an increase, so the log sheet, kept at a stated load, is the instrument.
Air-cooled split systems use the same word for a different quantity: liquid line temperature minus outdoor air temperature, per Lennox service literature, compared to that unit's own charging chart. That approach is on the superheat and subcooling calculator; this page is the water-cooled chiller's.
Frequently Asked Questions
What is the chiller condenser approach formula?
Condenser approach is the saturated condensing temperature minus the leaving condenser water temperature. Trane writes it as "(the difference between the condensing refrigerant temperature and the leaving condenser water temperature)" and YORK as "(condenser temperature minus leaving condenser water temperature)". On the same machine at the same load and the same condenser water flow, a larger number is worse; across loads it is not, which is the rest of this page. The saturated condensing temperature is not measured with a thermometer: it is read from the condenser pressure on the refrigerant's PT table, which is the step this tool does for R-410A, R-134a, R-404A, R-32, R-22 and R-290. YORK's controller does the same from its transducer: "Check the condenser saturation temperature (based upon condenser pressure sensed by the condenser transducer) on the System Screen." On such a machine, enter the figure the screen shows.
What is a normal or good condenser approach temperature?
There is no number for the approach itself, and this tool prints none. Trane says fouling is indicated when the approach "is higher than predicted"; YORK says "Readings taken when a system is newly installed establish normal conditions with which to compare later readings." and, in its daily inspection for the YK, that the difference "must not exceed the difference recorded for a new unit by more than 4°F (2.2°C)." Both compare to the machine's own record. YORK's 4 F is an allowance over that record, for YORK's YK, not a value for the approach, and the arithmetic says why no such value can exist: At constant UA and water flow the approach is proportional to load, so a condenser that has lost 30 percent of its heat transfer can read a lower approach at half load than a clean one reads at full load, and a fixed threshold, or an unscaled comparison to the record, passes it. The comparisons that mean something are to your own commissioning record: scaled to the load the machine is carrying now, and by YORK's unscaled rule, both of which this tool makes once you enter the record.
Why does the tool refuse to say whether my approach is high?
Because without the machine's own record it cannot know, and guessing is how a fouled condenser gets called healthy. Effectiveness-NTU for a condenser with one isothermal fluid gives approach = Q / (m cp) x exp(-NTU) / (1 - exp(-NTU)): with UA and water flow held, the approach is proportional to load. Under this page's illustration (one assumed machine, labelled as such) a condenser at 70 percent of its clean UA reads 9.05 F at half load, lower than the 11.19 F the clean condenser reads at full load. A rule that says "above design approach means fouled" passes that machine. So without a record the tool prints the approach and stops. With a record and no present load it prints YORK's unscaled rule as YORK's and withholds the load-scaled comparison, because comparing to an unscaled record is the same trap. And if the record scaled to the present load is not a figure any condenser reads, the tool says so and withholds the scaled comparison rather than grade against it.
YORK's YK inspection allows 4 F over the recorded difference. Why does the tool also scale the record to load?
Because the two answer different questions and can disagree, and the page shows both. YORK's daily inspection for the YK reads "Check for any signs of dirty or fouled condenser tubes. The temperature difference between water leaving condenser and saturated condensing temperature must not exceed the difference recorded for a new unit by more than 4°F (2.2°C)." That is YORK's rule for YORK's machine, and the tool applies it as printed, unscaled. The load-scaled comparison is this site's, from the algebra: at 50 percent load a record taken at full load scales to half, so a reading that sits within YORK's 4 F can still be well above what this machine's record projects to at this load. When they disagree the result says so and why. Neither is applied to a machine YORK's manual does not cover as if it were that machine's rule; the scope rides in the sentence.
Which refrigerants can this tool convert from gauge pressure?
R-410A, R-134a, R-404A, R-32, R-22 and R-290: the fluids with a single-column PT table on this site. R-407C, R-454B, R-407A, R-448A and R-449A are zeotropic blends with two temperatures at one pressure (bubble and dew), so their pressure is not turned into one saturated temperature here; read the saturated condensing temperature the controller displays. R-507A, R-513A, R-450A, R-454A, R-454C, R-452B, R-123, R-1234yf, R-1234ze(E), R-515B, R-1233zd(E), R-514A, R-1224yd(Z), R-1336mzz(Z), R-600a, R-717, R-744 and R-1270 have no PT table on this site and are entered the same way. A pressure outside a table is refused rather than clamped to the table's edge.
My approach is above the scaled record. Does that mean the tubes are fouled?
Not yet. Two things raise the approach with clean tubes, and both need ruling out first: air in the condenser (non-condensables), which raises the condensing pressure and so the saturated temperature read from it, and low condenser water flow, which raises the approach at the same load. Neither can hide a fouled condenser, because both push the approach up, not down. On a low-pressure machine (R-123 and R-514A) the purge pump-out time is the check for air, and this page hands you to the purge tool with the refrigerant on the link. YORK's sentence is "For example, an increase in condenser approach temperature (condenser temperature minus leaving condenser water temperature) can be an indication of dirty condenser tubes." Can be, not is.
Is this the same approach temperature as on an air-cooled split system?
Air-cooled split systems use the same word for a different quantity: liquid line temperature minus outdoor air temperature, per Lennox service literature, compared to that unit's own charging chart. That approach is on the superheat and subcooling calculator; this page is the water-cooled chiller's.
Does the approach change after an R-134a to R-513A conversion?
Johnson Controls' SI0389 has the technician establish the baseline performance on R-134a before the conversion, and after it expects the discharge superheat and approach temperatures to differ from the R-134a values, while noting the approach temperatures may be similar. Either way the commissioning record on the old fluid is not the record for the new one: take a new baseline after the conversion, at a known load, and compare to that. Trane's sentence, for what a rise means on any machine: "Condenser tube fouling is indicated when the approach temperature (the difference between the condensing refrigerant temperature and the leaving condenser water temperature) is higher than predicted."
Related Calculators
Chiller Purge Pump-Out Leak Check
Air in the condenser mimics fouling. For R-123 and R-514A CenTraVac, purge minutes per day against Trane's Figure 2 by tons per circuit.
R-134a Chiller Retrofit Eligibility
Which refrigerants EPA lists for retrofit in an R-134a centrifugal chiller, and what the OEM bulletin requires; the approach record is retaken after a conversion.
Superheat & Subcooling Calculator
The DX world's approach (liquid line minus outdoor air) lives here, with PT interpolation for 11 refrigerants.
Approach up on a low-pressure machine? Check the purge first.
Air in the condenser raises the condensing pressure, and the approach with it, on clean tubes. For R-123 and R-514A CenTraVac the purge pump-out time is read against Trane's own chart by tonnage.