TSC 3900 DPAnalysis Rev E

Dew/Frost Point Uncertainty Analysis of the
Model 3900
Two-Temperature, Two-Pressure Low Humidity Generator
Revision E
September 28, 2017
1 Introduction
Described here is the Dew Point and Frost Point Uncertainty Analysis, following NIST Guideline 1297 1,
for a Model 3900 Humidity Generator, manufactured by Thunder Scientific Corporation, that combines the
NIST developed and proven two-temperature and two-pressure humidity generation principles.2,3
Generating gas of a known dew point or frost point temperature in a system of this type does not require
direct measurements of the water vapor content of the gas. Rather, the generated dew point and/or frost
point temperature is derived from the measurements of saturation temperature, saturation pressure, and the
pressure at the point of use, commonly referred to as either test pressure or chamber pressure. For the
purposes of this analysis, the terms ‘test pressure’ and ‘chamber pressure’ are synonymous with each other.
2 Defining Equations
2.1 Common Equations
The following equations of Wexler/Hardy7 for saturation vapor pressure, enhancement factor, and
temperature (from saturation vapor pressure) are common and fundamental to most humidity calculations
presented here.
2.1.1 Saturation Vapor Pressure over Water, e

Saturation vapor pressure over water at a given ITS-90 temperature is defined by the formula [7]
 6 
e  exp   g i T i 2  g 7 ln T  (1)
 i 0 
where e is the saturation vapor pressure, in Pascals, over liquid water in the pure phase,
T is the temperature in Kelvin,
and g0 = -2.8365744·103
g1 = -6.028076559·103
g2 = 1.954263612·101
g3 = -2.737830188·10-2
g4 = 1.6261698·10-5
g5 = 7.0229056·10-10
g6 = -1.8680009·10-13
g7 = 2.7150305
2.1.2 Saturation Vapor Pressure over Ice, e

Saturation vapor pressure over ice at a given temperature is defined by the formula [7]
Uncertainty Analysis of the Thunder Scientific Model 3900 “Two-Pressure Two-Temperature” Low Humidity Generator Page 1 of 29
Copyright © 2016, Thunder Scientific Corporation. All Rights Reserved.
Document : TSC 3900 DPAnalysis.doc Rev E
 4 
e  exp   k i T i 1  k 5 ln T  (2)
 i 0 
where e is the saturation vapor pressure, in Pascals, over ice in the pure phase
T is the temperature in Kelvin
and k0 = -5.8666426·103
k1 = 2.232870244·101
k2 = 1.39387003·10-2
k3 = -3.4262402·10-5
k4 = 2.7040955·10-8
k5 = 6.7063522·10-1
2.1.3 Enhancement Factors

The ‘effective’ saturation vapor pressure over water or ice in the presence of other gases differs from the
ideal saturation vapor pressures given in equations 1 and 2. The effective saturation vapor pressure is
related to the ideal by
é = e·ƒ (3)
where é is the ‘effective’ saturation vapor pressure

e is the ideal saturation vapor pressure (as given in equation 1 or 2)
and ƒ is the enhancement factor.
The enhancement factor, for an air-water vapor mixture, is determined at a given temperature and pressure
from the formula [7]
  e  P 
ƒ  exp  1       1 (4)
  P e 
3
with    ai T i (5)
i 0
 3

and   exp   bi T i  (6)
 i 0 
where ƒ is the enhancement factor

e is the ideal saturation vapor pressure (as given in equation 1 or 2)
P is pressure in the same units as e
T is temperature in Kelvin
and aI, bI depend on temperature range and are given as
for water
223.15 to 273.15 K (-50 to 0 °C) 273.15 to 373.15 K (0 to 100 °C)
-2
a0 = -5.5898101·10 a0 = -1.6302041·10-1
a1 = 6.7140389·10-4 a1 = 1.8071570·10-3
a2 = -2.7492721·10-6 a2 = -6.7703064·10-6
a3 = 3.8268958·10-9 a3 = 8.5813609·10-9
b0 = -8.1985393·101 b0 = -5.9890467·101
b1 = 5.8230823·10-1 b1 = 3.4378043·10-1
b2 = -1.6340527·10-3 b2 = -7.7326396·10-4
b3 = 1.6725084·10-6 b3 = 6.3405286·10-7
for ice
173.15 to 223.15 K (-100 to -50 °C) 223.15 to 273.15 K (-50 to 0 °C)
-2
a0 = -7.4712663·10 a0 = -7.1044201·10-2
a1 = 9.5972907·10-4 a1 = 8.6786223·10-4
a2 = -4.1935419·10-6 a2 = -3.5912529·10-6
a3 = 6.2038841·10-9 a3 = 5.0194210·10-9
b0 = -1.0385289·102 b0 = -8.2308868·101
b1 = 8.5753626·10-1 b1 = 5.6519110·10-1
b2 = -2.8578612·10-3 b2 = -1.5304505·10-3
b3 = 3.5499292·10-6 b3 = 1.5395086·10-6
2.1.4 Temperature from Saturation Vapor Pressure

Equations 1 and 2 are easily solved for saturation vapor pressure over water or ice for a given saturation
temperature. However, if vapor pressure is known while temperature is the unknown desired quantity, the
solution immediately becomes complicated and must be solved by iteration. For ease of computation, the
following inverse equation is provided. This equation is generally used to find the dew point or frost point
temperature when the vapor pressure of a gas has been determined. When vapor pressure is known, use the
water coefficients to obtain dew point, and use the ice coefficients to obtain frost point.
 c ln e
i
i
T i 0
3
(7)
 d ln e
i
i
i 0
where T is the temperature in Kelvin

and e is the saturation vapor pressure in Pascals
with coefficients
for water for ice
c0 = 2.0798233·102 c0 = 2.1257969·102
c1 = -2.0156028·101 c1 = -1.0264612·101
c2 = 4.6778925·10-1 c2 = 1.4354796·10-1
c3 = -9.2288067·10-6 c3 = 0
d0 = 1 d0 = 1
d1 = -1.3319669·10-1 d1 = -8.2871619·10-2
d2 = 5.6577518·10-3 d2 = 2.3540411·10-3
d3 = -7.5172865·10-5 d3 = -2.4363951·10-5
2.2 Dew Point and Frost Point Determination

2.2.1 Definitions of Terms
Ts Saturation Temperature. The temperature at which the gas is fully saturated with water
vapor, and is most often made by a direct measurement of the temperature of the saturator
itself. Ts is in Kelvin, ts is in °C.
Tc Chamber Temperature. The temperature of the gas in the test chamber, or in the device
under test, at the location of the humidity sensor. Tc is in Kelvin, tc is in °C.
Td Dew Point Temperature. The temperature to which a gas must be cooled in order to just
begin condensing in the form of liquid dew. While contrary to common sense, liquid
dew can form in a meta-stable state at temperatures below freezing (called super-cooled
dew). Thus, dew point temperatures below 0 °C are quite common and reproducible.
While dew point and frost point exhibit identical vapor pressures, dew point and frost
point temperatures are not the same, except at 0.01 °C which is the triple point of water.
Td is in Kelvin; td is in °C.
Tf Frost Point Temperature. The temperature to which a gas must be cooled in order to just
begin condensing in the form of frost or ice. Frost point only exists at temperatures
below freezing (tf <= 0.01 °C). While frost point and dew point exhibit identical vapor
pressures, frost point and dew point temperatures are not the same, except at 0.01 °C
which is the triple point of water. Tf is in Kelvin, tf is in °C.
Ps Saturator Pressure. (As it applies to the humidity generators described in this document,
Saturation Pressure is synonymous with Saturator Pressure.) The total pressure in the
saturator, measured at the final point of saturation (generally the saturator outlet). This is
an absolute (not gauge) measurement. Ps is in Pascals.
Pc Chamber Pressure. The total pressure as measured in the test chamber, or at the device
under test, at the location of the humidity sensor. This is also referred to throughout this
document as Test Pressure. This is an absolute (not gauge) measurement. Pc is in
Pascals.
es Saturation Vapor Pressure at the Saturation Temperature. The partial pressure of the
water vapor in the saturator, as determined by measurement of the saturation temperature.
Regardless of the total pressure of the saturator, es is dependent on saturation temperature
only and further assumes that full saturation is actually being achieved. For temperatures
above freezing, es is computed as Saturation Vapor pressure over Water. For
temperatures below freezing, es is generally computed as Saturation Vapor Pressure over
Ice. es is expressed in Pascals.
ec Saturation Vapor Pressure at the Chamber Temperature. The maximum possible partial
pressure of water vapor that could exist in the test chamber, if the gas were fully saturated
with water vapor at the chamber temperature. Regardless of the total pressure of the
chamber, ec is dependent on chamber temperature only. For temperatures above freezing,
ec is computed as Saturation Vapor pressure over Water. For temperatures below
freezing, ec is generally computed as Saturation Vapor Pressure over Ice. However,
when using ec in the computation of %RH, but only when doing so in accordance with
the World Meteorogical Organization (WMO) adopted quidelines, ec is to be computed
with respect to water for all temperature conditions, even those below freezing. ec is
expressed in Pascals.
ed Saturation Vapor Pressure at the Dew/Frost Point Temperature (also known as Dew
Point Vapor Pressure, Frost Point Vapor Pressure, and partial water vapor pressure).
The partial pressure of the water vapor at the dew point temperature and computed with
respect to liquid water, or at the frost point temperature and computed with respect to ice.
Where frost point exists (at all temperatures below freezing), dew point vapor pressure
calculated at the dew point temperature with repect to liquid water and frost point vapor
pressure calculated at the frost point temperature with respect to ice are always equal and
synonymous terms. ed is expressed in Pascals.
s Enhancement Factor at Saturation Temperature and Saturation Pressure. The
enhancement factor corrects for the slight non-ideal behavior of water vapor when
admixed with other gases. The effective saturation vapor pressure that results under
saturation at any given temperature and pressure condition is determined by computing
the product of the saturation vapor pressure and the enhancement factor.
c Enhancement Factor at Chamber Temperature and Chamber Pressure. The
admixed with other gases. The effective saturation vapor pressure that results under
the product of the saturation vapor pressure and the enhancement factor.
d Enhancement Factor at Dew/Frost Point Temperature and Chamber Pressure. The
admixed with other gases. The effective dew/frost point vapor pressure that results under
the product of the dew/frost point vapor pressure and the enhancement factor.
2.2.2 Dew Point Temperature, Td

Dew point temperature is the temperature to which a gas must be cooled in order to just begin condensing
water vapor in the form of dew. Dew point temperature in this system is obtained with the following
iterative steps.
A. The vapor pressure es at the saturation temperature is calculated with equation 1 or equation 2. To
get es from this equation, the saturation temperature Ts is used for T. When Ts > 0, equation 1 is
used. When Ts < 0, specific knowledge of the state of the water (whether liquid or ice) is needed.
While it is possible for the water in a saturator to remain liquid for a short time when below 0 °C,
liquid water in a saturator operating below 0 °C will eventually freeze into a state of ice. (A
saturator that has been operating at or below –5 °C for more than an hour is most often expected to
be operating in a state of ice.) Once frozen into ice, the water remains in that state as long as the
saturation temperature remains below 0 °C. When the saturator is operating in a state of ice,
equation 2 is required.
B. The enhancement factor fs at the saturation temperature and saturation pressure is calculated using
equation 4. To get fs from this equation, calculations are performed using e = es, P = Ps, and T =
Ts. Equation 4 must be used with the correct coefficients (relative to the correct temperature range
for water or ice) based upon saturation temperature Ts and specific knowledge of the state of the
water in the saturator.
C. An educated guess is made at the dew/frost point enhancement factor, d. Setting d = 1 is a
suitable first guess.
D. The dew/frost point vapor pressure ed of the gas is computed with the two-pressure, two-
temperature relationship
ƒ s Pc
ed  e s   (8)
ƒ d Ps
E. Dew point temperature Td is calculated from dew/frost point vapor pressure ed using equation 7
and the coefficients for water. To get Td from this equation, calculations are performed using e =
ed.
F. The dew/frost point enhancement factor fd is calculated using equation 4 and the coefficients for
water of the appropriate range (based on the value of Td). To get fd from this equation,
calculations are performed using e = ed, P = Pc, and T = Td.
G. The dew point temperature, Td, converges to the proper value by iterating steps D through F
several times as necessary.
2.2.3 Frost Point Temperature, Tf
Frost point temperature is the temperature to which a gas must be cooled in order to just begin condensing
water vapor in the form of frost or ice. Frost point only exists at temperatures below freezing.
Note that generating a frost point does not require ice in the saturator, nor does generating a dew point temperature
require liquid water in the saturator. While the state of the saturator is required for proper application of the
equations and their calculations, the generator is ultimately controlling at some specified vapor pressure. Dew point
temperature and frost point temperature share the same vapor pressure and enhancement factor, but have different
numeric values of temperature. All vapor pressures that correspond to a dew point temperature at or below 0.01 °C
also have a corresponding frost point temperature. At a dew point temperature of 0.01 °C, frost point temperature is
equal to dew point temperature. For all values below that down to -50 °C, the two diverge from each other with dew
point temperature always lower in value than the corresponding frost point temperature. Dew point temperature and
frost point temperature, while different from each other in numeric value, are equally valid methods of expressing the
same vapor pressure. Dew point temperatures below -50 °C are considered unobtainable as no vapor pressure data
for water is available for temperature below -50 °C, while frost point temperature remains valid down to -100 °C as
vapor pressure date with respect to ice is available.
Frost point is obtained with the following iterative steps.
A. The vapor pressure es at the saturation temperature is calculated with equation 1 or equation 2. To
get es from this equation, the saturation temperature Ts is used for T. When Ts > 0, equation 1 is
used. When Ts < 0, specific knowledge of the state of the water (whether liquid or ice) is needed.
While it is possible for the water in a saturator to remain liquid for a short time when below 0 °C,
liquid water in a saturator operating below 0 °C will eventually freeze into a state of ice. (A
saturator that has been operating at or below –5 °C for more than an hour is most often expected to
be operating in a state of ice.) Once frozen into ice, the water remains in that state as long as the
saturation temperature remains below 0 °C. When the saturator is operating in a state of ice,
equation 2 is required.
B. The enhancement factor fs at the saturation temperature and saturation pressure is calculated using
equation 4. To get fs from this equation, calculations are performed using e = es, P = Ps, and T =
Ts. Equation 4 must be used with the correct coefficients (relative to the correct temperature range
for water or ice) based upon saturation temperature Ts and specific knowledge of the state of the
water in the saturator.
C. An educated guess is made at the dew/frost point enhancement factor, d. Setting d = 1 is a
suitable first guess.
D. The dew/frost point vapor pressure ed of the gas is using equation 8.
E. Frost point temperature Tf is calculated from dew/frost point vapor pressure ed using equation 7
and the coefficients for ice. To get Tf from this equation, calculations are performed using e = ed.
F. The dew/frost point enhancement factor fd is calculated using equation 4 and the coefficients for
ice of the appropriate range (based on the value of Tf). To get fd from this equation, calculations
are performed using e = ed, P = Pc, and T = Tf.
G. The frost point temperature, Tf, converges to the proper value by iterating steps D through F
several times as necessary.
3 Uncertainty
To analyze the overall expanded uncertainty in generated dew point and/or frost point temperature, the
uncertainties associated with temperatures and pressures must be determined, along with other possible
sources of uncertainty. These individual components of uncertainty must then be statistically combined to
form the Combined Uncertainty. The Expanded Uncertainty is then determined by multiplying the
Combined Uncertainty by a suitable coverage factor, k, based on the desired confidence level.
Due to the complexity associated with the computations, and the iterative requirement to reach a final
solution, algebraic methods involving partial derivatives of the underlying equations prove difficult.
Rather, a more straightforward approach will be taken that utilizes a table of sensitivity coefficients at
various temperature and pressure combinations. The table will identify the sensitivity of generated dew
and frost point temperature to uncertainty in the saturation temperature, saturation pressure, and test
pressure. Construction of the sensitivity tables will then allow straight forward determination of the
uncertainty in dew and frost point temperature due to:
• uncertainty in saturation pressure which includes

measurement uncertainty
measurement hysteresis
measurement resolution
• uncertainty in test pressure which includes
• uncertainty in saturation temperature which includes
• uncertainty contribution from saturator efficiency
• uncertainty in vapor pressure
• uncertainty in enhancement factors
• uncertainty contribution due to adsorption, desorption, and permeation
3.1 Creation of Sensitivity Coefficients at Various Saturation

Temperatures and Pressures
Due to the complexity in developing algebraic solutions, a computer science approach is taken to determine
these values. The calculation steps of sections 2.2.2 and 2.2.3 along with equations 1 through 8 are
programmed into a computer that determines dew point and frost point temperatures using variable inputs
of saturation temperature Ts, saturation pressure Ps, and test pressure Pc. A sensitivity coefficient is then
determined by calculating the first derivative of the dew/frost point equation for each of the given inputs of
saturation temperature, saturation pressure, and chamber pressure at various combinations.
Table 1. Dew Point & Frost Point Sensitivity Coefficients

Humidity Generator Conditions Frost Point Sensitivity Coefficients Dew Point Sensitivity Coefficients
Nominal
FP Ts Ps ΔFP/ΔTs ΔFP/ΔPs ΔFP/ΔPc ΔDP/ΔTs ΔDP/ΔPs ΔDP/ΔPc
[°C] [°C] [kPa] [°C/°C] [°C/kPa] [°C/kPa] [°C/°C] [°C/kPa] [°C/kPa]
-95 -80 1668.93 0.843 0.003 0.051
-79.05 2000 0.834 0.002
-80 597.029 0.897 0.009
-90 -75 1414.74 0.849 0.003
-73.04 2000 0.829 0.002 0.054
-80 101.325 1.000 0.060
-72.37 344.74 0.925 0.017
-80
-70 498.099 0.903 0.012
-60.91 2000 0.821 0.003 0.060
-70 101.325 1.000 0.066
-61.54 344.74 0.921 0.019
-70 -60 427.266 0.907 0.015
-50 1667.27 0.823 0.004
-48.66 2000 0.812 0.003 0.066
-60 101.325 1.000 0.072
-50.66 344.74 0.917 0.021
-60 -50 374.63 0.912 0.019
-40 1277.76 0.832 0.005
-36.28 2000 0.803 0.003 0.072
-50 101.325 1.000 0.079
-40 334.27 0.915 0.024
-50 -39.73 344.74 0.913 0.023
-30 1020.07 0.839 0.008
-23.76 2000 0.794 0.004 0.079
-40 101.325 1.000 0.087 1.057 0.092
-30 302.6 0.919 0.029 0.971 0.030
-40 -28.76 344.74 0.909 0.025 0.961 0.027
-20 839.95 0.846 0.010 0.894 0.011
-11.1 2000 0.785 0.004 0.087 0.830 0.004 0.092
-30 101.325 1.000 0.094 1.074 0.101
-20 277.21 0.922 0.034 0.990 0.037
-17.73 344.74 0.905 0.027 0.973 0.030
-30
-10 708.82 0.851 0.013 0.915 0.014
0 1723.92 0.738 0.005 0.793 0.006
1.96 2000 0.682 0.004 0.094 0.732 0.005 0.101
-20 101.325 1.000 0.102 1.093 0.112
-10 256.5 0.925 0.040 1.010 0.044
-6.66 344.74 0.901 0.030 0.985 0.033
-20
0 610.31 0.805 0.017 0.879 0.018
10 1248.98 0.695 0.008 0.760 0.009
16.91 2000 0.657 0.005 0.102 0.718 0.005 0.112
-10 101.325 1.000 0.111 1.112 0.123
0 239.37 0.872 0.047 0.970 0.052
-10 5.08 344.74 0.784 0.032 0.873 0.036
10 484.44 0.754 0.023 0.838 0.025
17 770.59 0.713 0.014 0.111 0.793 0.016 0.123
0 101.325 1.000 0.135
0 10 204.24 0.922 0.067
17 323.42 0.872 0.042 0.135
10 101.325 1.000 0.147
10
17 160.19 0.946 0.093 0.147
Notice that in the table of sensitivity coefficients, nominal frost and dew point values are shown with a
variety of saturation temperature and pressure combinations. For each specific frost point or dew point
temperature listed, an attempt was made to include the maximum possible saturation pressure, minimum
saturation pressure, and a pressure equal to the switch-over point between the high and low range saturation
pressure transducer. In addition, the lowest and highest possible saturation temperature was also listed.
For all of the above calculations, a common test pressure (standard atmospheric pressure) is assumed.
3.2 Uncertainty Contribution from Pressure
Determining the uncertainty in generated output based on saturation pressure and test pressure requires
knowledge of uncertainty in the pressure measurement, pressure hysteresis, and measurement resolution of
the saturator and test pressure transducers.
3.2.1 Pressure Measurement

This system utilizes three absolute pressure transducers. One is used for the test pressure measurement.
While it has a full scale of 0 to 345 kPa absolute, it is generally used only for the barometric pressure range.
For the purposes of this analysis, standard barometric pressure of 101.325 kPa is assumed. The other two
pressure transducers are of different ranges and are used for the measurement of saturation pressure. For
low saturation pressures (those below 345 kPa), the low range transducer is used. For higher saturation
pressure, where the uncertainty in saturation pressure is of a lesser concern, a higher range transducer is
used.
Pressure measurement uncertainty was analyzed from the data collected during annual calibration of the
pressure transducers. During calibration, each transducer was tested at no fewer than 3 points over its
specific range using a total system calibration approach. With this approach, the transducers remain
electrically connected to the system allowing the pressure transducer, the measuring electronics, and the
displayed data to be calibrated as a complete system rather than as individual components. Data gathered
during the calibration is system rather than component data. The combined data from several years of
calibration history were used in the computation of statistical standard deviations. For each of the pressure
transducers (which includes the measurement electronics and display), the standard deviation, P, from the
desired mean values were determined to be:
P = 0.069 kPa (0.01 psia) for P<345 kPa (<50 psia)

P = 0.276 kPa (0.04 psia) for P>345 kPa (>50 psia)
Using normal distribution, the pressure uncertainties normalized to one sigma are equivalent to the standard
deviation values given above for each of the transducers.
 uPc[meas] = 0.069 kPa (0.01 psia)
uPs[meas] = 0.069 kPa (0.01 psia) for P<345 kPa (<50 psia)
= 0.276 kPa (0.04 psia) for P>345 kPa (>50 psia)
The statistical standard deviations calculated from the calibration history also have an uncertainty
component from the Mensor PCS400 pressure standard used during the calibration process. The uncertainty
of the Mensor PCS400 pressure standard for the low and high-pressure ranges are as follows:
uPc[std] = 0.046 kPa
uPs[std] = 0.046 kPa for P<345 kPa (<50 psia)

= 0.228 kPa for P>345 kPa (>50 psia)
3.2.2 Pressure Hysteresis
The system most often operates for hours, days, or weeks on end at one saturation pressure. When a
change in saturation pressure is made, the new condition is again maintained for hours, days, or weeks.
The test pressure transducer generally monitors changes only in the barometric pressure. Since both the
saturator and test pressure measurements result in very slow moving, nearly steady state conditions,
hysteresis in pressure measurement is nearly imperceptible. However, the affect of any hysteresis must still
be considered in the overall analysis. Based on rectangular distribution, hysteresis in pressure
measurement normalized to one sigma, is estimated to be
 uPc[hyst] = 0.035 /3 kPa (0.005/3 psia)

= 0.020 kPa (0.003 psia)
uPs[hyst] = 0.035 /3 kPa (0.005/3 psia) for P<345 kPa (50 psia)
= 0.020 kPa (0.003 psia) for P<345 kPa (50 psia)
= 0.172/3 kPa (0.025/3 psia) for P>345 kPa (50 psia)

= 0.099 kPa (0.014 psia) for P>345 kPa (50 psia)
3.2.3 Pressure Resolution

The analog to digital conversion process resolves 1 part in 25000 over the range of each of the pressure
transducers. The resolution is then computed as
resolutionp = (TransducerRange)/25000
The resolution for each of the transducers is then
resolutionpc = 345 kPa / 25000

= 0.014 kPa (0.002 psia)
resolutionps = 345 kPa / 25000

= 0.014 kPa for P<345 kPa (0.002 psia for P<50 psia)
= 2068 kPa / 25000

= 0.083 kPa for P>345 kPa (0.012 psia for P>50 psia)
Based on a rectangular distribution of the half-interval of resolution, the uncertainty of pressure due to
resolution normalized to one sigma is computed by multiplying the resolution by 0.5/3. The uncertainty
of pressure due to resolution is then
uP[res] = resolutionp * 0.5/3
uPc[res] = 0.004 kPa (0.0006 psia)
uPs[res] = 0.004 kPa for P<345 kPa (0.0006 psia for P<50 psia)
= 0.024 kPa for P>345 kPa (0.0035 psia for P>50 psia)
3.2.4 Summary of Pressure Uncertainties
Pressure uncertainties of various types may be combined statistically using the following
(uP)2 = (uP[meas])2 + (uP[std])2 + (uP[hyst])2 + (uP[res])2 …
Uncertainty data in tables 2 through 4 are combined in that manner.
Table 2. Test Pressure Uncertainty, uPc

Source Type Uncertainty
[kPa]
uP c[meas] A 0.069
uP c[std] B 0.023
uP c[hyst] B 0.020
uP c[res] B 0.004
uP c 0.076
Table 3. Saturation Pressure Uncertainty, uPs (for Ps<345 kPa)

[kPa]
uP s[meas] A 0.069
uP s[std] B 0.023
uP s[hyst] B 0.020
uP s[res] B 0.004
uP s 0.076
(P s < 345 kPa)
Table 4. Saturation Pressure Uncertainty, uPs (for Ps>345 kPa)

[kPa]
uP s[meas] A 0.276
uP s[std] B 0.077
uP s[hyst] B 0.099
uP s[res] B 0.024
uP s 0.304
(P s > 345 kPa)
3.3 Uncertainty Contribution from Temperature
Determining the uncertainty in generated output based on saturation temperature requires knowledge of
uncertainty in the measurement, the resolution, and the self-heating of the saturation thermometer.
Saturation temperature is measured by a 1k ohm thermistor to measure the saturation temperature range of
–80 to +17 °C. Temperature measurement uncertainty was analyzed from the data collected during annual
calibration of the saturator temperature probe. During calibration, the saturation thermometer was tested at
no fewer than 3 points over its measurement range using a total system calibration approach. With this
approach, the thermometer remains electrically connected to the system allowing the temperature probe, the
measuring electronics, and the displayed data to be calibrated as a complete system rather than as
individual components. Data gathered during the calibration is system rather than component data. The
combined data from several years of calibration history were used in the computation of statistical standard
deviations. The standard deviation of the saturation thermometer system, T, from the desired mean
values were determined to be:
T [meas] = 0.023 °C
Using normal distribution, the temperature measurement uncertainty normalized to one sigma is equivalent
to the standard deviation value given above for the saturation thermometer.
 uTs[meas] = 0.023 °C
The statistical standard deviations calculated from the calibration history also have an uncertainty
component from the Hart 1560 Black Stack & 5626 PRT temperature standard used during the calibration
process. The uncertainty of the Hart 1560 Black Stack & 5626 PRT temperature standard is as follows:
uTs[std] = 0.006 °C
3.3.1 Temperature Resolution

The analog to digital conversion process which transforms the saturation thermometer resistance into
digital values resolves to 0.002 °C. Based on a rectangular distribution of the half-interval, the uncertainty
component of saturation temperature resolution is then
uTs[res] = 0.01 * 0.5/3

= 0.003
3.3.3 Self Heating

The saturation temperature probe is generally calibrated and checked in a well-stirred fluid bath. In use, it
is also immersed directly within a pumped fluid surrounding the saturator. Since the conditions of
calibration and use are very similar, both immersed within moving fluid, the self heating is considered
almost negligible. However, an estimate of uncertainty will be applied.
 uTs[self heating] = 0.005 °C
3.3.4 Thermal Lag
The saturator is of a stacked plate design, constructed completely of stainless steel, sealed and immersed in
a pumped fluid medium. Direction of the fluid flow is counter to that of the saturator gas stream.
The temperature of the pumped fluid medium is controlled to the desired saturation temperature and
measured by the saturation temperature probe. Given adequate time, the saturator outlet is assumed to
come into thermal equilibrium with the average temperature of the pumped fluid medium. However,
during times of temperature transition, the saturator plates will lag the temperature of the fluid by up to
several degrees. No attempt will be made here to predict the uncertainty associated with thermal lag.
However it will be assumed that adequate time is allowed for the saturator to regain thermal equilibrium
with the pumped fluid medium prior to relying on the data from the generator. Lag times of 30 minutes to
1 hour are not considered uncommon. When approaching the final value, the rate of change is very slow
and becomes difficult to detect on the instrument under test. Therefore, an estimate of uncertainty will be
applied.
 uTs[thermal lag] = 0.01 °C
3.3.5 Thermal Gradients

Design of the saturator is that of a counter-flow design where the fluid medium flows in a direction
opposite that of the gas stream being saturated. Thermal gradients do exist within the saturator from inlet
to outlet. Controlling the direction of this gradient is important to proper saturation. The temperature of
the fluid is measured and controlled at the point it enters the saturator cavity, which is the same point that
the saturated gas stream exits the saturator due to the counter flow design. The temperature will be slightly
higher at the fluid exit point, which is also the gas entry point. Provided the saturator is of sufficient
thermal capacity and effective path length, complete thermal transfer between the gas flowing in one
direction and the fluid flowing in the opposite direction will ensure that the exiting gas has reached thermal
equilibrium with the entering fluid and is therefore at fluid temperature. An estimate of uncertainty will be
applied.
 uTs[gradient] = 0.005 °C
[It is believed that the design of the saturator reduces any negative affects that a temperature gradient might otherwise cause if
uncontrolled or improperly directed. Furthermore, it is believed that this design actually improves the ability of the saturator to fully
saturate the gas stream with water vapor, thereby improving saturator efficiency.]
3.3.6 Temperature Control Stability

Temperature stability relates to the ability of the temperature control system to maintain a constant
temperature in the pumped fluid medium, and ultimately in the saturator itself. The saturator temperature
control system maintains the saturation temperature fluid at the setpoint with a standard deviation of 0.02
°C. Using normal distribution, the uncertainty in temperature relating to control stability normalized to one
sigma is equivalent to the standard deviation. It is given therefore as
 uTs[stability] = 0.02 °C
3.3.7 Summary of Temperature Uncertainties
Temperature uncertainties of various types may be combined statistically using the following
(uT)2 = (uT[meas])2 + (uT[resolution])2 + (uT[std])2 …
Uncertainty data in table 5 is combined in that manner.
Table 5. Saturation Temperature Uncertainty, uT

[°C]
uT s[meas] A 0.023
uT s[res] B 0.003
uT s[std] B 0.006
uT s[self heating] B 0.005
uT s[thermal lag] B 0.010
uT s[gradients] B 0.005
uT s[control stability] B 0.020
uT s 0.033
3.4 Saturation Efficiency,

All two-pressure two-temperature humidity generators of single pass design rely on the ability of the
saturator to fully saturate the gas with water vapor as it passes from inlet to outlet. Based on the counter
flow design of the saturator (fully discussed in Thermal Gradients), it is assumed for all practical purposes
that the saturator efficiency is 100%. Even given that assumption, small differences in saturation within the
system can lead to uncertainty in the generated dew and frost point temperatures. This uncertainty is very
small and very difficult to isolate or even measure outright, making it difficult to accurately quantify its
value. Because of this, an instrument comparison was used to help identify this uncertainty. The premise of
the instrument comparison was an RH Systems 373 chilled mirror. Data gathered during instrument
comparison tests from several years of calibration history resulted in an estimated 99.95% efficiency.
Table 6 shows this estimated efficiency in terms of dew and frost point temperature uncertainty.
Table 6. Saturation Efficiency Uncertainty, uSE

Nominal FP uSE
[°C] [°C]
-95 0.003
-90 0.003
-80 0.003
-70 0.003
-60 0.004
-50 0.004
-40 0.004
-30 0.005
-20 0.005
-10 0.006
0 0.006
10 0.008
3.5 Vapor Pressure

Vapor pressure equations are given in equations 1 and 2 for water and ice respectively. There are
uncertainties assigned to the experimental data used as a basis for those equations. When calculating dew
or frost point temperature from the humidity generator, the vapor pressure equations are only used as a
means to transform data from saturation temperature to vapor pressure, then from vapor pressure back to
dew or frost point temperature. Since the vapor pressure equations are used simply as bi-directional
transfer functions, and round-trip transformation is always made, uncertainty in the data used to create
these functions is of little significance, especially when the saturation temperature and dew or frost
temperature are nearly equal. Under that circumstance, the dew and frost point vapor pressure is nearly
equal to the saturation vapor pressure. However, when the ratio of saturation temperature to dew or frost
temperature is high, the dew and frost point vapor pressure is lower than the saturation vapor pressure
(likewise, the dew or frost point temperature is lower than the saturation temperature). This results in the
possibility that the transformation function in one direction (temperature to saturation vapor pressure)
occurs on a different portion of the function’s curve than the inverse transformation (vapor pressure to dew
or frost point temperature). Due to slight variations of the original experimental data, the quality of fit of
the function to the data, and other factors, some allowance must be made for uncertainty associated with
use of the vapor pressure equations. Table 7 shows this estimated vapor pressure uncertainty in terms of
dew and frost point temperature.
Table 7. Vapor Pressure Uncertainty, uSE

Nominal FP Ts Ps uVp
[°C] [°C] [kPa] [°C]
-80 1669.85 0.028
-95
-79.05 2000.00 0.028
-80 597.27 0.029
-90 -75 1414.74 0.027
-73.04 2000.00 0.027
-80 101.33 0.030
-72.37 344.74 0.029
-80
-70 498.10 0.028
-60.91 2000.00 0.026
-70 101.33 0.029
-61.54 344.74 0.027
-70 -60 427.27 0.026
-50 1667.30 0.024
-48.66 2000.00 0.024
-60 101.33 0.027
-50.66 344.74 0.025
-60 -50 374.63 0.024
-40 1277.80 0.023
-36.28 2000.00 0.022
-50 101.33 0.025
-40 334.27 0.022
-50 -39.73 344.74 0.022
-30 1020.10 0.021
-23.76 2000.00 0.019
-40 -40 101.33 0.022
-30 302.60 0.020
-28.76 344.74 0.020
-20 839.95 0.018
-11.1 2000.00 0.016
-30 101.33 0.019
-20 277.21 0.017
-17.73 344.74 0.016
-30
-10 708.82 0.015
0 1723.90 0.014
1.96 2000.00 0.014
-20 101.33 0.015
-10 256.50 0.012
-6.66 344.74 0.011
-20
0 610.31 0.011
10 1249.00 0.011
16.91 2000.00 0.011
-10 101.33 0.009
0 239.37 0.006
-10 5.08 344.74 0.006
10 484.44 0.006
17 770.59 0.006
0 101.33 0.001
0 10 204.24 0.001
17 323.42 0.001
10 101.33 0.001
10
17 160.17 0.002
3.6 Enhancement Factor

Enhancement factors are slight correction factors used to account for the non-ideal behavior of water vapor
when admixed with other gases. The enhancement factor is dependent on both temperature and pressure
and is given in equation 4. In use, the enhancement factor is applied as a multiplier to the vapor pressure
resulting in the ‘effective vapor pressure’. For analytical purposes the enhancement factor can be
considered in a manner similar to the vapor pressure for determining the affect of uncertainty in the
enhancement factor data and formula. In the case of enhancement factor the pressure differences between
saturator and test pressure results in the possibility that the transformation function in one direction occurs
on a different portion of the function’s curve than the inverse transformation. Table 8 shows this estimated
enhancement factor uncertainty in terms of dew and frost point temperature.
Table 8. Enhancement Factor Uncertainty, uEF

Nominal FP Ts Ps uEF
[°C] [°C] [kPa] [°C]
-80 1669.85 0.045
-95
-79.05 2000.00 0.054
-80 597.27 0.016
-90 -75 1414.74 0.037
-73.04 2000.00 0.052
-80 -80 101.33 0.004
-72.37 344.74 0.010
-70 498.10 0.013
-60.91 2000.00 0.047
-70 101.33 0.004
-61.54 344.74 0.009
-70 -60 427.27 0.010
-50 1667.30 0.036
-48.66 2000 0.043
-60 101.33 0.004
-50.66 344.74 0.008
-60 -50 374.63 0.009
-40 1277.80 0.026
-36.28 2000.00 0.040
-50 101.33 0.004
-40 334.27 0.008
-50 -39.73 344.74 0.008
-30 1020.10 0.020
-23.76 2000.00 0.037
-40 101.33 0.004
-30 302.60 0.007
-40 -28.76 344.74 0.008
-20 839.95 0.016
-11.1 2000.00 0.034
-30 101.33 0.003
-20 277.21 0.006
-17.73 344.74 0.007
-30
-10 708.82 0.013
0 1723.90 0.016
1.96 2000.00 0.019
-20 101.33 0.003
-10 256.50 0.005
-6.66 344.74 0.006
-20
0 610.31 0.007
10 1249.00 0.011
16.91 2000.00 0.017
-10 101.33 0.002
0 239.37 0.003
-10 5.08 344.74 0.004
10 484.44 0.005
17 770.59 0.007
0 101.33 0.002
0 10 204.24 0.003
17 323.42 0.004
10 101.33 0.002
10
17 160.17 0.003
3.7 Permeation, Adsorption and Desorption
Permeation, adsorption and desorption refers to a continuous influx from or outgas to the humidity of the
surrounding environment (such as the air within the laboratory) through small leaks or semi-permeable
surfaces through the walls, fittings, valves, and dead spaces within the system. Although somewhat
difficult to accurately quantify, instrument comparison has shown that these permeation affects start to
become noticeable at frost point temperatures below approximately –60 °C with the greatest negative
impact at the lowest frost point values. In the case of this generator, permeation affects are due to
permeation of water vapor from the high ambient room conditions, through the tubing, fittings, or valves,
into the dry gas stream output of the generator. Permeation tends to increase the frost point temperature of
the generated gas stream. It is estimated that permeation tends to increase the concentration of water vapor
in the gas stream independent of saturation temperature, saturation pressure, and test pressure. However,
the effect of this permeation is most noticeable when running at the lowest frost points (where the
concentration of water vapor in the gas is also very low), and at low flow rates. By increasing the flow rate
of the generated gas, while assuming a constant permeation rate, the affect of that permeation is minimized.
Like wise, if the gas being generated is of higher concentration (i.e., warmer frost point), then the affect of
added water vapor from permeation is reduced. While the nominal frost point value being generated may
be very low, it is recommended that the highest flow rate possible be used for the low frost point conditions
to minimize the affect of permeation. This analysis assumes that the generator is run at the highest possible
flow rate. Even with the high flow rate, permeations impact the generation of low frost point values. This
permeation uncertainty is again very small and very difficult to isolate and measure outright, making it
difficult to accurately quantify. Because of this, the same instrument comparison described in section 3.4
was also used to identify the uncertainty caused by permeation. Permeation leads more to an uncorrected
bias than to a random variation in output, resulting in an offset during the instrument comparison tests.
This offset was estimated to be five parts per billion (ppb). Table 9 shows the estimated permeation
uncertainty bias in terms of frost point temperature.
Table 9. Permeation, absorption and desorption uncertainty bias, ubias

Nominal bias
[°C] [°C FP]
-95 0.736
-90 0.292
-80 0.056
-70 0.013
-60 0.003
-50 0.001
-40 0.000
Because permeation leads more to an uncorrected bias than to a random variation in output it will be treated
slightly different in the analysis. Uncertainty associated with uncorrected bias will be algebraically added to
the expanded uncertainty rather than statistically included in combined uncertainty.
4. Combined Standard Uncertainty

The combined standard uncertainty, uc, is obtained by statistical combination of the individual uncertainty
components of pressure, temperature, and others, each multiplied by their associated sensitivity
coefficients. Statistical combination is performed in accordance with the following:
uc2 = uc12 + uc22 + … + ucn2
where uc1, uc2, … ucn are individual components of uncertainty each

multiplied by their respective sensitivity coefficients
The following tables reflect the standard uncertainty components and the combined standard uncertainty at
various frost and dew point temperatures at a variety of saturation pressures and temperatures.
Table 10. Combined Uncertainty at 10°C Dew Point

Combined Uncertainty at 10°C Dew Point
Ts = 10.00 Ts = 17.00
Ps = 101.33 Ps = 160.19
Assigned Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.147 0.011 0.147 0.011
Sat Pressure <345 0.076 0.147 0.011 0.093 0.007
uP s
Sat Pressure >345 0.304
Sat Temp uT s 0.033 1.000 0.033 0.946 0.032
Vapor Pressure uVp 0.001 0.002
Enhancement Factor uEF 0.002 0.003
Saturation Efficiency
uSE 0.008 0.008
combined uc 0.038 0.035
Table 11. Combined Uncertainty at 0°C Dew Point

Combined Uncertainty at 0°C Dew Point
Ts = 0.00 Ts = 10.00 Ts = 17.00
Ps = 101.33 Ps = 204.24 Ps = 323.42

Assigned Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.135 0.010 0.135 0.010 0.135 0.010
Sat Pressure <345 0.076 0.135 0.010 0.067 0.005 0.042 0.003
uP s
Sat Pressure >345 0.304
Sat Temp uT s 0.033 1.000 0.033 0.922 0.031 0.872 0.029
Vapor Pressure uVp 0.001 0.001 0.001
Enhancement Factor uEF 0.002 0.003 0.004
uSE 0.006 0.006 0.006
combined uc 0.037 0.033 0.032
Table 12. Combined Uncertainty at -10°C Frost Point
Combined Uncertainty at -10°C Frost Point
Ts = -10.00 Ts = 0.00 Ts = 5.08 Ts = 5.08 Ts = 10.00 Ts = 17.00
Ps = 101.33 Ps = 239.37 Ps = 344.74 Ps = 344.74 Ps = 484.44 Ps = 770.59
Assigned Sens Std Sens Std Sens Std Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.111 0.008 0.111 0.008 0.111 0.008 0.111 0.008 0.111 0.008 0.111 0.008
Sat Pressure <345 0.076 0.111 0.008 0.047 0.004 0.032 0.002
uP s
Sat Pressure >345 0.304 0.032 0.010 0.023 0.007 0.014 0.004
Sat Temp uT s 0.033 1.000 0.033 0.872 0.029 0.784 0.026 0.784 0.026 0.754 0.025 0.713 0.024
Vapor Pressure uVp 0.009 0.006 0.006 0.006 0.006 0.006
Enhancement Factor uEF 0.002 0.003 0.004 0.004 0.005 0.007
uSE 0.006 0.006 0.006 0.006 0.006 0.006
combined uc 0.037 0.032 0.029 0.031 0.029 0.028

Ts = -20.00 Ts = -10.00 Ts = -6.66 Ts = -6.66 Ts = 0.00 Ts = 10.00 Ts = 16.91
Ps = 101.33 Ps = 256.50 Ps = 344.74 Ps = 344.74 Ps = 610.31 Ps = 1249.0 Ps = 2000.0
Assigned Sens Std Sens Std Sens Std Sens Std Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.102 0.008 0.102 0.008 0.102 0.008 0.102 0.008 0.102 0.008 0.102 0.008 0.102 0.008
Sat Pressure <345 0.076 0.102 0.008 0.040 0.003 0.030 0.002
uP s
Sat Pressure >345 0.304 0.030 0.009 0.017 0.005 0.008 0.002 0.005 0.001
Sat Temp uT s 0.033 1.000 0.033 0.925 0.031 0.901 0.030 0.901 0.030 0.805 0.027 0.695 0.023 0.657 0.022
Vapor Pressure uVp 0.015 0.012 0.011 0.011 0.011 0.011 0.011
Enhancement Factor uEF 0.003 0.005 0.006 0.006 0.007 0.011 0.017
uSE 0.005 0.005 0.005 0.005 0.005 0.005 0.005
combined uc 0.039 0.035 0.034 0.035 0.031 0.030 0.031
Ts = -30.00 Ts = -20.00 Ts = -17.73 Ts = -17.73 Ts = -10.00 Ts = 0.00 Ts = 1.96
Ps = 101.33 Ps = 277.21 Ps = 344.74 Ps = 344.74 Ps = 708.82 Ps = 1723.9 Ps = 2000.0
Assigned Sens Std Sens Std Sens Std Sens Std Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.094 0.007 0.094 0.007 0.094 0.007 0.094 0.007 0.094 0.007 0.094 0.007 0.094 0.007
Sat Pressure <345 0.076 0.094 0.007 0.034 0.003 0.027 0.002
uP s
Sat Pressure >345 0.304 0.027 0.008 0.013 0.004 0.005 0.002 0.004 0.001
Sat Temp uT s 0.033 1.000 0.033 0.922 0.031 0.905 0.030 0.905 0.030 0.851 0.028 0.738 0.025 0.682 0.023
Vapor Pressure uVp 0.019 0.017 0.016 0.016 0.015 0.014 0.014
Enhancement Factor uEF 0.003 0.006 0.007 0.007 0.013 0.016 0.019
uSE 0.005 0.005 0.005 0.005 0.005 0.005 0.005
combined uc 0.040 0.037 0.036 0.037 0.036 0.034 0.033

Ts = -40.00 Ts = -30.00 Ts = -28.76 Ts = -28.76 Ts = -20.00 Ts = -11.10
Ps = 101.33 Ps = 302.60 Ps = 344.74 Ps = 344.74 Ps = 839.95 Ps = 2000.0
Sat Pressure <345 0.076 0.087 0.007 0.029 0.002 0.025 0.002
uP s
Sat Pressure >345 0.304 0.025 0.008 0.010 0.003 0.004 0.001
Sat Temp uT s 0.033 1.000 0.033 0.919 0.031 0.909 0.030 0.909 0.030 0.846 0.028 0.785 0.026
uSE 0.004 0.004 0.004 0.004 0.004 0.004
combined uc 0.041 0.038 0.038 0.038 0.038 0.046
Ts = -50.00 Ts = -40.00 Ts = -39.73 Ts = -39.73 Ts = -30.00 Ts = -23.76
Ps = 101.33 Ps = 334.27 Ps = 344.74 Ps = 344.74 Ps = 1020.1 Ps = 2000.0
Sat Pressure <345 0.076 0.079 0.006 0.024 0.002 0.023 0.002
uP s
Sat Pressure >345 0.304 0.023 0.007 0.008 0.002 0.004 0.001
Sat Temp uT s 0.033 1.000 0.033 0.915 0.031 0.913 0.030 0.913 0.030 0.839 0.028 0.794 0.026
uSE 0.004 0.004 0.004 0.004 0.004 0.004
combined uc 0.043 0.039 0.039 0.040 0.041 0.050

Ts = -60.00 Ts = -50.66 Ts = -50.66 Ts = -50.00 Ts = -40.00 Ts = -36.28
Ps = 101.33 Ps = 344.74 Ps = 344.74 Ps = 374.63 Ps = 1277.8 Ps = 2000.0
Sat Pressure <345 0.076 0.072 0.005 0.021 0.002
uP s
Sat Pressure >345 0.304 0.021 0.006 0.019 0.006 0.005 0.002 0.003 0.001
Sat Temp uT s 0.033 1.000 0.033 0.917 0.031 0.917 0.031 0.912 0.030 0.832 0.028 0.803 0.027
uSE 0.004 0.004 0.004 0.004 0.004 0.004
combined uc 0.044 0.041 0.041 0.041 0.045 0.053
Ts = -70.00 Ts = -61.54 Ts = -61.54 Ts = -60.00 Ts = -50.00 Ts = -48.66
Ps = 101.33 Ps = 344.74 Ps = 344.74 Ps = 427.27 Ps = 1667.3 Ps = 2000.0
Sat Pressure <345 0.076 0.066 0.005 0.019 0.001
uP s
Sat Pressure >345 0.304 0.019 0.006 0.015 0.005 0.004 0.001 0.003 0.001
Sat Temp uT s 0.033 1.000 0.033 0.921 0.031 0.921 0.031 0.907 0.030 0.823 0.027 0.812 0.027
uSE 0.003 0.003 0.003 0.003 0.003 0.003
combined uc 0.045 0.042 0.042 0.042 0.052 0.056

Ts = -80.00 Ts = -72.37 Ts = -72.37 Ts = -70.00 Ts = -60.91
Ps = 101.33 Ps = 344.74 Ps = 344.74 Ps = 498.10 Ps = 2000.0
Assigned Sens Std Sens Std Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.060 0.005 0.060 0.005 0.060 0.005 0.060 0.005 0.060 0.005
Sat Pressure <345 0.076 0.060 0.005 0.017 0.001
uP s
Sat Pressure >345 0.304 0.017 0.005 0.012 0.004 0.003 0.001
Sat Temp uT s 0.033 1.000 0.033 0.925 0.031 0.925 0.031 0.903 0.030 0.821 0.027
Vapor Pressure uVp 0.030 0.029 0.029 0.028 0.026
Enhancement Factor uEF 0.004 0.010 0.010 0.013 0.047
uSE 0.003 0.003 0.003 0.003 0.003
combined uc 0.046 0.043 0.044 0.043 0.060
Ts = -80.00 Ts = -75.00 Ts = -73.04
Ps = 597.03 Ps = 1414.7 Ps = 2000.0
Assigned Sens Std Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.054 0.004 0.054 0.004 0.054 0.004
Sat Pressure <345 0.076
uP s
Sat Pressure >345 0.304 0.009 0.003 0.003 0.001 0.002 0.001
Sat Temp uT s 0.033 0.897 0.030 0.849 0.028 0.829 0.028
Vapor Pressure uVp 0.029 0.027 0.027
Enhancement Factor uEF 0.016 0.037 0.052
uSE 0.003 0.003 0.003
combined uc 0.045 0.054 0.064

Ts = -80.00 Ts = -79.05
Ps = 1668.9 Ps = 2000.0
Assigned Sens Std Sens Std
Source Name Uncertainty Coef Uncert Coef Uncert
Test Pressure uP c 0.076 0.051 0.004 0.051 0.004
Sat Pressure <345 0.076
uP s
Sat Pressure >345 0.304 0.003 0.001 0.002 0.001
Sat Temp uT s 0.033 0.843 0.028 0.834 0.028
Vapor Pressure uVp 0.028 0.028
Enhancement Factor uEF 0.045 0.054
uSE 0.003 0.003
combined uc 0.060 0.067
5 Expanded Uncertainty
Utilizing a coverage factor k=2, the expanded uncertainty, U, is computed using the formula
U = (k * uc ) + bias
Where k = the coverage factor (2 for 95% confidence level)

uc = uncertainties at specific saturation temperatures and saturation pressures. uc
is obtained from the tables listed in section 4.
bias = any uncorrected bias associated with the nominal frost point or at a specific
saturation temperature and saturation pressure. Values of bias are obtained
from the table in section 3.7
Like the combined uncertainties listed in the tables of section 4, there are two separate calculations of
expanded uncertainty at the saturation pressure of 344.74 kPa for many of the nominal frost point values.
This saturation pressure is at the switch point between use of the low-pressure transducer and use of the
high-pressure transducer. At this saturation pressure, either transducer may be in use. If the low range
transducer is in use, the associated expanded uncertainty is lower. If the high range transducer has
switched in, then the expanded uncertainty will be larger in value. There is a step change (or more
precisely, a discontinuity) in the associated expanded uncertainty at this possible pressure switching point.
The expanded uncertainty values are listed in table 22. Expanded uncertainty values are shown in °C frost
point (°C dew point for nominal generated values above 0 °C).
Table 22. Expanded Uncertainty ( k=2)

Nominal
FP Ts Ps
[°C] [°C] [kPa] U
-80 1668.93 0.857
-95
-79.05 2000 0.870
-80 597.029 0.382
-90 -75 1414.74 0.400
-73.04 2000 0.421
-80 101.325 0.148
-72.37 344.74 0.143
-80 -72.37 344.74 0.144
-70 498.099 0.143
-60.91 2000 0.177
-70 101.325 0.102
-61.54 344.74 0.097
-61.54 344.74 0.098
-70
-60 427.266 0.097
-50 1667.27 0.117
-48.66 2000 0.126
-60 101.325 0.091
-50.66 344.74 0.085
-50.66 344.74 0.086
-60
-50 374.63 0.085
-40 1277.76 0.093
-36.28 2000 0.110
-50 101.325 0.086
-40 334.27 0.080
-39.73 344.74 0.080
-50
-39.73 344.74 0.081
-30 1020.07 0.082
-23.76 2000 0.101
-40 101.325 0.083
-30 302.6 0.076
-28.76 344.74 0.076
-40
-28.76 344.74 0.077
-20 839.95 0.076
-11.1 2000 0.093
-30 101.325 0.080
-20 277.21 0.073
-17.73 344.74 0.072
-30 -17.73 344.74 0.074
-10 708.82 0.071
0 1723.92 0.067
1.96 2000 0.067
-20 101.325 0.078
-10 256.5 0.071
-6.66 344.74 0.069
-20
-6.66 344.74 0.071
0 610.31 0.064
10 1248.98 0.060
16.91 2000 0.063
-10 101.325 0.075
0 239.37 0.065
5.08 344.74 0.059
-10
5.08 344.74 0.062
10 484.44 0.059
17 770.59 0.057
0 101.325 0.075
0 10 204.24 0.068
17 323.42 0.065
10 101.325 0.076
10
17 160.19 0.071
Note: 10 °C nominal frost point listed above is actually 10 °C dew point. Expanded uncertainties at this nominal dew point are also
listed in °C dew point.
Graph 1: The following graph indicates the expanded uncertainties at various saturation pressures. Each
nominal frost/dew point value is shown separately.
Expanded Uncertainty as a function of Saturation Temperature
1.000
10°C DP
0.900
0°C DP
0.800
DP & FP Uncertainty [°C]
-10°C FP
0.700 -20°C FP
0.600 -30°C FP
-40°C FP
0.500
-50°C FP
0.400 -60°C FP
0.300 -70°C FP
0.200 -80°C FP
-90°C FP
0.100
-95°C FP
0.000
-100 -80 -60 -40 -20 0 20 40
Saturation Temperature [°C]
Graph 2: The following graph indicates the expanded uncertainties at various saturation temperatures.
Expanded Uncertainty as a function of Saturation Pressure
1.000
10°C DP
0.900
0°C DP
0.800
-10°C FP
0.700 -20°C FP
0.600 -30°C FP
-40°C FP
0.500
-50°C FP
0.400 -60°C FP
0.300 -70°C FP
0.200 -80°C FP
Uncertainty Analysis of the Thunder Scientific Model 3900 “Two-Pressure Two-Temperature” Low Humidity Generator Page 26-90°C
of 29 FP
0.100
Document : TSC 3900 DPAnalysis.doc Rev E -95°C FP
0.000
0 500 1000 1500 2000 2500
Saturation Pressure [kPa]
Each nominal frost/dew point value is shown separately.
Note that at each frost/dew point value, there is only small variation in the uncertainty at the various
saturation temperature and saturation pressure combinations. In order to obtain uncertainty as a direct
function of frost/dew point temperature, independent of the saturation temperature and pressure
combinations, the average uncertainty value at each nominal frost/dew point value is chosen. The result of
using the rounded average value of expanded uncertainty at each nominal frost/dew point is illustrated in
the following table.
Table 23. Maximum Expanded Uncertainty (k=2)

Nominal FP
[°C] U
-95 0.9
-90 0.5
-80 0.2
-70 0.1
-60 0.1
-50 0.1
-40 0.1
-30 0.1
-20 0.1
-10 0.1
0 0.1
10 0.1
Graph 3: This graph, from the above table, depicts the maximum uncertainty at each of the nominal
frost/dew point temperatures.
Expanded Uncertainty as a function of Frost/Dew Point Temperature
1.000
0.900
0.800
0.700
0.600
0.500
0.400
0.300
0.200
0.100
0.000
-95 -90 -85 -80 -75 -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10
Frost/Dew Point Temperature [°C]
6 Summary
Expanded uncertainty of the low humidity generator in terms of dew point and frost point is relatively
constant between +10 and about –75 °C, varying only gradually over that entire range. For values below
approximately –75 °C, the affect of permeation tends to dominate the uncertainty.
It is also worth noting that the uncertainty at any given frost or dew point temperature is relatively constant
regardless of the specific combinations of saturation temperature and saturation pressure. This allows the
system to adequately generate frost or dew point temperatures without much regard for the specific
saturation temperature-pressure combination chosen.
It is also noted that there was no assumed permeation, adsorption and desorption assigned to the RH
Systems 373 chilled mirror comparison instrument. In reality the 373 has some quantifiable bias of
permeation, adsorption and desorption for the very same reasons the 3900 does.
References:
1. Taylor, Barry N. and Kuyatt, Chris E., Guidelines for Evaluating and Expressing the Uncertainty of
NIST Measurement Results, NIST Technical Note 1297, 1994 Edition
2. Wexler, Arnold and Daniels, Raymond, Pressure-Humidity Apparatus, Journal of Research of the
National Bureau of Standards, April 1952, Vol. 48, No. 4, 269-274.
3. Hasagawa, S. and Little, J.W., The NBS Two-Pressure Humidity Generator, Mark 2, Journal of
Research of the National Bureau of Standards – A. Physics and Chemistry, January-February 1977,
Vol. 81A, No. 1, 81-88
4. Wexler, Arnold, Vapor Pressure Formulations for Water in Range 0 to 100 C. A Revision., Journal of
Research of the National Bureau of Standards - A. Physics and Chemistry, September-December 1976,
Vol. 80A, Nos. 5 and 6, 775-785, Equation 15.
5. Greenspan, L., Functional Equations for the Enhancement Factors for CO 2-Free Moist Air, Journal of
Research of the National Bureau of Standards – A. Physics and Chemistry, January-February 1976,
Vol. 80A, No.1, 41-44
6. Kuyatt, Chris, et al., Determining and Reporting Measurement Uncertainties, Recommended Practice
RP-12, National Conference of Standards Laboratories, April 1995
7. Hardy, Bob, ITS-90 Formulations for Vapor Pressure, Frostpoint Temperature, Dewpoint
Temperature, and Enhancement Factors in the range -100 to +100C, Proceedings of the Third
International Symposium on Humidity and Moisture, Teddington, London, England, April 1998
8. Barry N. Taylor and Chris E. Kuyatt, “Guidelines for Evaluating and Expressing the Uncertainty of
NIST Measurement Results”, NIST Technical Note 1297, 1994 Edition
9. A. Wexler and R.W. Hyland, “In Thermodynamic properties of dry air, moist air and water and SI
psychrometric charts”, 1983 ASHRAE (Project 216−RP), Table 20.

TSC 3900 DPAnalysis Rev E

Uploaded by

Copyright:

Available Formats

TSC 3900 DPAnalysis Rev E

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TSC 3900 DPAnalysis Rev E

Uploaded by

Copyright:

Available Formats

Dew/Frost Point Uncertainty Analysis of the

2.1.1 Saturation Vapor Pressure over Water, e

2.1.2 Saturation Vapor Pressure over Ice, e

2.1.3 Enhancement Factors

where é is the ‘effective’ saturation vapor pressure

where ƒ is the enhancement factor

2.1.4 Temperature from Saturation Vapor Pressure

where T is the temperature in Kelvin

2.2 Dew Point and Frost Point Determination

2.2.2 Dew Point Temperature, Td

Frost point is obtained with the following iterative steps.

• uncertainty in saturation pressure which includes

3.1 Creation of Sensitivity Coefficients at Various Saturation

Table 1. Dew Point & Frost Point Sensitivity Coefficients

3.2.1 Pressure Measurement

P = 0.069 kPa (0.01 psia) for P<345 kPa (<50 psia)

 uPc[meas] = 0.069 kPa (0.01 psia)

uPc[std] = 0.046 kPa

uPs[std] = 0.046 kPa for P<345 kPa (<50 psia)

 uPc[hyst] = 0.035 /3 kPa (0.005/3 psia)

= 0.172/3 kPa (0.025/3 psia) for P>345 kPa (50 psia)

3.2.3 Pressure Resolution

The resolution for each of the transducers is then

resolutionpc = 345 kPa / 25000

resolutionps = 345 kPa / 25000

= 2068 kPa / 25000

uP[res] = resolutionp * 0.5/3

uPc[res] = 0.004 kPa (0.0006 psia)

(uP)2 = (uP[meas])2 + (uP[std])2 + (uP[hyst])2 + (uP[res])2 …

Uncertainty data in tables 2 through 4 are combined in that manner.

Table 2. Test Pressure Uncertainty, uPc

Table 3. Saturation Pressure Uncertainty, uPs (for Ps<345 kPa)

Table 4. Saturation Pressure Uncertainty, uPs (for Ps>345 kPa)

3.3.1 Temperature Resolution

uTs[res] = 0.01 * 0.5/3

3.3.3 Self Heating

 uTs[self heating] = 0.005 °C

 uTs[thermal lag] = 0.01 °C

3.3.5 Thermal Gradients

3.3.6 Temperature Control Stability

(uT)2 = (uT[meas])2 + (uT[resolution])2 + (uT[std])2 …

Uncertainty data in table 5 is combined in that manner.

Table 5. Saturation Temperature Uncertainty, uT

3.4 Saturation Efficiency,

Table 6. Saturation Efficiency Uncertainty, uSE

3.5 Vapor Pressure

Table 7. Vapor Pressure Uncertainty, uSE

3.6 Enhancement Factor

Table 8. Enhancement Factor Uncertainty, uEF

Table 9. Permeation, absorption and desorption uncertainty bias, ubias

4. Combined Standard Uncertainty

uc2 = uc12 + uc22 + … + ucn2

where uc1, uc2, … ucn are individual components of uncertainty each

Table 10. Combined Uncertainty at 10°C Dew Point

combined uc 0.038 0.035

Table 11. Combined Uncertainty at 0°C Dew Point

Ps = 101.33 Ps = 204.24 Ps = 323.42

combined uc 0.037 0.033 0.032

combined uc 0.037 0.032 0.029 0.031 0.029 0.028