Skip to content
Industrial Unit Converter

Engineering Guide

Steam Quality and Dryness Fraction: IAPWS-IF97 Formula and Throttling Calorimeter Method

Published July 3, 2026 · by Industrial Unit Converter Editorial Team

Steam Quality and Dryness Fraction: IAPWS-IF97 Formula and Throttling Calorimeter Method

A pharmaceutical plant outside Boston ran 270°F sterilization cycles in four autoclaves. Spore tests kept failing. The BMS logs showed 15-minute dwell times, correct temperatures, and saturated steam at 50 psig from the clean steam generator. The shift supervisor pulled a Spirax Sarco throttling calorimeter, throttled the line to atmospheric pressure, and read 245°F. Two minutes of arithmetic gave a dryness fraction of 0.92, meaning 8% of the mass entering each autoclave was liquid water. The water was sitting on the load, displacing dry steam contact. The fix was a failed separator baffle upstream of the generator. The cost was about $9,400 in scrapped production batches over the six weeks it took to find.

Steam quality, the dryness fraction x, is the mass fraction of saturated vapor in a wet-steam mixture. x = 0 is saturated liquid, x = 1.0 is dry saturated steam, and 0 < x < 1 is wet steam. Superheated steam is defined as x = 1.0, with the degree of superheat reported as T minus T_sat. Sterilization, food cookers, paper drying, refinery tracing, turbine staging: all assume a dryness fraction they are not actually getting.

This article covers the formula, the IAPWS-IF97 reference tables, the throttling calorimeter measurement, two worked examples with real arithmetic, and the field mistakes that keep showing up in audits. The distribution side, headers, condensate legs, and trap selection, is in our Steam System Design: Headers, Condensate, and Traps guide.

The steam quality formula

For wet steam, the specific enthalpy of the mixture is the mass-weighted sum of the saturated liquid and saturated vapor enthalpies:

h = h_f + x * h_fg = h_f + x * (h_g - h_f)

Solving for x:

x = (h_actual - h_f) / h_fg = (h_actual - h_f) / (h_g - h_f)

The terms:

  • h_actual: specific enthalpy of the wet steam, BTU/lb
  • h_f: specific enthalpy of saturated liquid at the line pressure, BTU/lb
  • h_g: specific enthalpy of saturated vapor at the line pressure, BTU/lb
  • h_fg: latent heat of vaporization = h_g - h_f, BTU/lb
  • x: dryness fraction, 0 <= x <= 1

The procedure in any field or design problem is the same: measure h_actual, look up h_f and h_g at the local pressure, and divide. Most of the mistakes happen before the arithmetic, in the pressure reading or the sample extraction.

The formula is valid only for saturated and wet steam. For superheated steam, the temperature carries the extra information. x = 1.0 by definition, and the degree of superheat is the meaningful number. Our Steam Quality Calculator caps x at 1.0 and reports the superheat temperature above that.

IAPWS-IF97 reference data and the h_g peak

The IAPWS-IF97 Industrial Formulation (2012 release) is the international standard for steam properties. The values below are exact to the IAPWS-IF97 formulation, rounded for field use.

P (psig) P (psia) T_sat (degF) h_f (BTU/lb) h_g (BTU/lb) h_fg (BTU/lb)
0 14.7 212.0 180.2 1150.6 970.4
50 64.7 281.0 250.1 1174.0 923.9
100 114.7 338.0 298.6 1187.4 888.8
150 164.7 366.0 330.6 1194.0 863.4
200 214.7 388.0 355.4 1197.0 841.6
250 264.7 406.0 376.1 1198.4 822.3
300 314.7 421.0 393.9 1199.0 805.1

The ASME Steam Tables (2014 compilation) and Keenan, Keyes, Hill, and Moore's Steam Tables (1969) give the same numbers to four significant figures.

One row in this table is worth staring at. h_g peaks around 250 to 300 psig at 1198 to 1199 BTU/lb, then drops at higher pressures. At 600 psig h_g is only 1199.0 BTU/lb and the latent heat has fallen to 727.4 BTU/lb. This is why reheat cycles in utility boilers run at moderate pressures, and why condensing turbines stage the largest enthalpy drop in the low-pressure section where h_fg is still useful. The old rule of thumb "high pressure, more work" stops being true above about 1000 psig for a single-stage turbine.

The deeper dive on the IAPWS-IF97 formulation itself, including the region definitions and the supplementary release for supercritical conditions, is in our Steam Tables IAPWS-IF97 Deep Dive.

Throttling calorimeter measurement

A throttling calorimeter takes a small sample of the wet steam, runs it through a calibrated orifice into a chamber at atmospheric pressure, and measures the temperature of the resulting superheated steam. The most common instruments in industrial service are Spirax Sarco's throttling calorimeters, the Gestra calorimeter series, and the older Yarway design still in use in plants built before 1990. Cost is $3,000 to $8,000 for a Spirax Sarco unit, $10,000 to $20,000 for a separating calorimeter. Accuracy on x: roughly +/- 0.5% for throttling mode, +/- 0.3% for separating mode.

The procedure is the same regardless of vendor:

  1. Extract the sample at a dedicated sample port within ten pipe diameters downstream of the last riser, drip, or PRV. The sample line should be short, sloped downward, and insulated. Long, uninsulated sample lines re-condense moisture and bias the reading high.
  2. Throttle through the calibrated orifice to atmospheric pressure (14.7 psia at sea level; correct for elevation above 3,000 ft).
  3. Wait for the downstream temperature to stabilize, typically 60 to 90 seconds.
  4. Read the temperature and the line pressure simultaneously from a gauge at the sample port.

The governing physics is the throttling equation. A throttle does no work, transfers no heat, and therefore conserves enthalpy: h_after = h_before. The downstream condition is superheated because the orifice dropped the pressure below saturation. With h_after from the superheat tables and h_before = h_after, the calculation is the same x formula applied to the line pressure.

A subtler issue. If the steam in the line is superheated (T_line > T_sat at P_line), the calorimeter still works, but the result will be x > 1.0. Reporting a quality in that case is wrong. Report the degree of superheat: T_line - T_sat at P_line. This is the rule that catches the most people on first audits.

Worked example: 50 psig line, calorimeter reads 14.7 psia at 250 degF

A 50 psig line serves a food processing cooker. A Spirax Sarco throttling calorimeter reads 14.7 psia downstream at 250 degF. The line pressure is 64.7 psia (50 psig plus atmospheric).

Step 1: Look up h_after from the superheat tables. At 14.7 psia, T_sat = 212 degF, and 250 degF is 38 degF of superheat. From IAPWS-IF97 superheat tables: h_after at 14.7 psia, 250 degF is 1,167.4 BTU/lb.

Step 2: Apply the throttling equation. h_actual at the line = h_after = 1,167.4 BTU/lb.

Step 3: Look up h_f and h_fg at 64.7 psia (50 psig). h_f = 250.1 BTU/lb, h_fg = 923.9 BTU/lb, h_g = 1,174.0 BTU/lb.

Step 4: Calculate x. x = (1,167.4 - 250.1) / 923.9 = 917.3 / 923.9 = 0.993.

The steam is 99.3% dry, or 0.7% moisture by mass. That is a passing result for process steam and within the 0.95 minimum most food-grade specs require. It is also a result worth reading carefully: h_after (1,167.4) is below h_g at 50 psig (1,174.0), so the original steam was saturated or very slightly wet, and the quality arithmetic applies. If h_after had been 1,176.0, the steam was superheated, and the right answer would have been "2 degF of superheat" rather than any dryness fraction.

Power plant case: 100 MW turbine exhaust quality

A 100 MW condensing turbine at a Midwest utility runs at 1,000 psig, 1,000 degF throttle conditions and exhausts to a 1 psia condenser. The exhaust quality is the limiting parameter for last-stage blade life.

Inlet. At 1,014.7 psia and 1,000 degF, the steam is superheated. From IAPWS-IF97 superheat tables: h_1 = 1,505 BTU/lb, s_1 = 1.6521 BTU/(lb-degR).

Isentropic expansion to 1 psia. s_2s = s_1 = 1.6521 BTU/(lb-degR). At 1 psia: s_f = 0.1326, s_g = 1.9776, s_fg = 1.8450.

Isentropic exit quality. x_2s = (s_2s - s_f) / s_fg = (1.6521 - 0.1326) / 1.8450 = 1.5195 / 1.8450 = 0.823.

So x_2s = 0.823, or 82.3% quality, in the ideal isentropic case.

The real turbine runs at a lower exhaust quality, not higher. Turbine irreversibilities (staging losses, leakage, windage) all increase entropy, which means a wetter expansion and a lower quality. A typical reheat unit with good maintenance runs 0.75 to 0.85 quality at the last-stage exit. A unit with worn nozzles or excessive leakage can drop below 0.75.

Below about 0.70 quality the last-stage blades see enough liquid water to cause erosion. Modern low-pressure turbines have moisture separator reheaters (MSRs) between the HP and LP sections that pull 10 to 15% of the flow out, separate the water, and reheat the rest before re-entry. On a 100 MW unit, an MSR typically adds 1.0 to 1.5 percentage points to cycle efficiency and roughly doubles last-stage blade life. The full efficiency chain is in our Boiler Efficiency: Direct vs Indirect Methods guide and the Boiler Efficiency Calculator.

Common mistakes

1. Treating the calorimeter result as a quality when the steam is superheated. If the line is superheated, the throttling arithmetic gives x > 1.0. The result is sometimes reported as "101% quality" by operators who do not understand what is happening. The correct response is to recognize that the steam is superheated, and the number to report is T_line - T_sat at P_line. A surprising number of DCS displays report "steam quality" on every cycle without checking the upstream state.

2. Sampling at the wrong location. Steam picks up moisture in drip legs and loses moisture through pressure-reducing stations. Sampling at the boiler outlet gives a different answer than sampling at the process use point. For a clean steam generator serving a sterilizer, the sample port must be within 10 diameters of the sterilizer inlet, and the line should rise continuously. Falling sections are a common root cause of failed audits.

3. Using gauge pressure instead of absolute in the table lookup. A 50 psig gauge is 64.7 psia absolute. Looking up h_f and h_g at "50 psig" in a table that expects absolute pressure gives h_f for 50 psia, which is at a much lower temperature and produces a wildly wrong quality. Always add 14.7 to the gauge reading before the table lookup.

4. Assuming the boiler is producing dry steam. Carryover from the boiler drum, especially on a high-pressure firetube unit cycling near rated capacity, can drop quality at the outlet to 0.95 or lower. In one Midwest food plant, the operators had assumed "saturated steam" meant x = 1.0 from the boiler. The first audit gave 0.94, traced to a faulty mechanical separator in the steam drum, costing about $60,000 per year in extra fuel.

5. Ignoring the altitude correction. At 5,000 ft elevation, atmospheric pressure is about 12.2 psia, not 14.7. The downstream saturation temperature is 191 degF, not 212 degF. Plant sites above 3,000 ft should have altitude-corrected steam tables on hand.

6. Confusing moisture and quality in a process spec. A spec that says "max 2% moisture" is the same as "min 0.98 quality." Putting both in the same document is a sign of an editor who has not been on a plant floor.

Standards and best practices

  • IAPWS-IF97 (2012 release), the international standard for industrial steam properties.
  • ASME Steam Tables (2014), the U.S. equivalent compiled from the IAPWS-IF97 formulation.
  • ASME PTC 4 (Fired Steam Generators, 2013), the performance test code for boiler efficiency, with explicit requirements on steam quality sampling and measurement uncertainty.
  • ASME BPVC Section I (2023), the power boiler construction code, with limits on steam quality leaving the boiler (typically >= 0.98) and on superheat temperature.
  • EN 12953-11 (2003), the European shell boiler acceptance test standard, used alongside ASME BPVC Section I for international projects.
  • Spirax Sarco Steam Tables (current edition) and the Armstrong Steam Reference Book are the two most-used field handbooks, with worked throttling calorimeter examples.

Frequently asked questions

What is acceptable steam quality for industrial use?

For general process heating, 0.95 minimum (5% moisture max). For sterilization and food contact, 0.98 minimum. For steam turbine inlets, 0.99 minimum, with the last-stage exhaust typically at 0.75 to 0.85 in operation. For HVAC clean steam humidification, 0.98. A quality below 0.90 is almost always a fault, indicating carryover, separator failure, or a leaking PRV dumping condensate into the line.

How do I measure steam quality in the field?

Throttling calorimeter is the default. Accuracy about +/- 0.5% on x. Separating calorimeter for higher accuracy (+/- 0.3%) and where a throttling unit reads "x > 1" (superheated source). Conductivity probe for continuous monitoring at lower cost. Raman spectroscopy is used in research; rarely seen in production plants.

What is the difference between saturated steam and superheated steam?

Saturated steam is at the boiling temperature for its pressure. Superheated steam is hotter than T_sat, with the difference being the degree of superheat. The dryness fraction is defined only for saturated and wet steam. For superheated steam, the relevant parameter is the degree of superheat, not a quality.

How does steam quality affect heat transfer?

Wet steam delivers less total heat per pound at the same pressure. Beyond that, liquid droplets wet the heat transfer surface and can cause film boiling, which has a much lower heat transfer coefficient than nucleate boiling. In food sterilization, the 8% moisture case from the opening of this article was failing because the wet film on the load was insulating it from the steam.

What is the h_g peak, and why does it matter?

The enthalpy of saturated vapor peaks at about 250 to 300 psig at 1,198 to 1,199 BTU/lb, then declines at higher pressures. Above about 2,200 psig, the latent heat is small and most of the work in a turbine comes from sensible heat. This is the reason most utility boilers reheat to moderate pressure, and the reason condensing turbines stage the largest pressure drop in the low-pressure section.

What is the difference between steam quality and steam purity?

Quality is a thermodynamic property: the fraction of mass that is vapor. Purity is a chemistry property: the fraction of the total mass that is water molecules. A clean steam generator running on deionized feedwater can have x = 0.95 and still be 99.99% pure water. A process steam line with rust and silica carryover can be x = 1.0 and have only 95% water. Both numbers matter, and they are measured with different instruments.

References

  • IAPWS-IF97 (2012 release). International Association for the Properties of Water and Steam. Release on the IAPWS Industrial Formulation 1997 for the Thermodynamic Properties of Water and Steam.
  • ASME Steam Tables (2014). American Society of Mechanical Engineers. Companion tables to the IAPWS-IF97 formulation.
  • ASME PTC 4 (2013). Fired Steam Generators Performance Test Code. American Society of Mechanical Engineers.
  • ASME BPVC Section I (2023). Rules for Construction of Power Boilers. American Society of Mechanical Engineers.
  • Keenan, Keyes, Hill, and Moore (1969). Steam Tables: International Edition (Metric Units). John Wiley & Sons. The 1969 edition; the data tables remain in print and in field use.
  • Spirax Sarco (current edition). The Steam and Condensate Loop Book. Spirax-Sarco Limited. Field handbook with worked calorimeter examples.
  • EN 12953-11 (2003). Shell Boilers, Part 11: Acceptance Tests. European Committee for Standardization.

Related tools and calculators

Share LinkedIn X Email