NOTICE CONCERNING COPYRIGHT

RESTRICTIONS

This document may contain copyrighted materials. These materials have

been made available for use in research, teaching, and private study, but
may not be used for any commercial purpose. Users may not otherwise
copy, reproduce, retransmit, distribute, publish, commercially exploit or
otherwise transfer any material.

The copyright law of the United States (Title 17, United States Code)
governs the making of photocopies or other reproductions of copyrighted
material.

Under certain conditions specified in the law, libraries and archives are
authorized to furnish a photocopy or other reproduction. One of these
specific conditions is that the photocopy or reproduction is not to be "used
for any purpose other than private study, scholarship, or research." If a
user makes a request for, or later uses, a photocopy or reproduction for
purposes in excess of "fair use," that user may be liable for copyright
infringement.

This institution reserves the right to refuse to accept a copying order if, in
its judgment, fulfillment of the order would involve violation of copyright
law.
 Geothermal Resources Council TRANSACTIONS, Vol. 19, October 1995

~ ~ o t ~ Steam
e r Quality
~ a ~ Testing

DouglasB. Jung

Two-Phase Engineering & Research, Inc.

3209 Franz Valley Road, Sauta Rosa, CA 95404

Abstract
Geothermal steam quality and purity have a coinmon and even if a process upset is detected,
s i ~ c ~effecti t on the operational efficiency and many hours to days can pass before a moderate upset
life of geothermal steam turbines and accessory ciui be c o ~ e andd corrective ~easurestaken.
equipment. Poor steam processing can result in Damage can often occur during this window period.
scaled nozzles / blades, erosion, corrosion, reduced
util~tion efliciency, and early fatigue failures Accurate geothennal steain quality and purity
accelerated by stress corrosion cracking (SCC). ~ e ~ u r e m e n tare
s i n ~ s i c ~complex.
y First a
Upsets formed by undetected slugs of liquid entering representative sample must be taken, which in itself
tlie turbine can cause catastrophic failure. is difficult, then that sample must be processed into
meatlingfiil results either by trace chemistry or
The accurate ~ o ~ t and o ~ dge t e ~ a t i o n of t h e r m o d ~ ~means.
c It is not ~ c o ~ too incur
n
geothermal steam quality / purity is intrinsically errors in excess of 1Dx if conditions and techniques
complex which often results in substantial errors. are not proper. In the measurement of effective steam
This paper will review steam quality and purity qualities, values of 99.9999% can, in fact, be less
relationships, address some of the errors, 'than 99% because of comp~e~ties or proper
complexities, calibration and focus on: technique. This false sense of security can lead to
thennodynamic techniques for evaluating and serious problems.
~ o ~ t o steam
~ i g quality by use of tlie modified
throttling calorimeters.
This paper will address some of tlie complexities
inlierent in the measurement of geothennal steam
-quality and purity. The review will include basic
'Intr o ~ u c t i o ~ relation~psbetween quality and purity; isokinetic
sampling, probe effects and multi-phase flow
Steam quality and purity entering geothermal plants regimes; trace chemistry, conductivity, and
are iderior to &el fired systems. This processed thermodyiamic analytical tools; and calibration.
steam contains gasses, solids and dissolved solids in
varying quantities. As a result, geothermal steam Although trace chemistry techniques can provide
turbines, etc., require higher maintenance aid incur better precision, the bulk of tliis paper will address
lggher premature failure rates tlm conventional &el the tlxottling calorimeter for the determination of
'fired steam plants. steam q ~ t yThis
. rather old and expensive tool, if
properly utilized can provide fast and accurate results
Uillike conventional steam plants that continuously without chemicals or expensive and delicate
monitor fluid chemistry, geothermal plants take kboratory equipment. This is an instrument that can
o c c ~ i o ngrab
~ samples, perhaps once per shift, and be used in the field. As with any ~ a l ~ ctool,
a l good
sample at individual site processing stations less judgment and cross calibration with other teclmiques
eequently. Sampling and analytical errors are is prudent.

575
Jung
Steam Quality Quality vs. Purity

Steam quality is the percentage by weight of dry steam in If the moisture in the steam is separated brine of a
a , mixture of saturated steam and suspended liquid known concentration, the following equality exists
droplets of water at the same temperature. Thus 99.90% between steam quality aid purity:
steam quality contains 0.10% moisture. Steam quality is
determined by thermodynamic means via use of the
throttling calorimeter on saturated steam samples. In Steam Quality = ( I - pm impuritv ) x 100 (I)
geothermal applications, heat loss, Joule-Thomson ppm separated brine
effects, barometric pressure, static/stagnation state, and
other effects must be corrected for meaningfid results.
ppm impurity = % moisture x PDM in sep. brine (2)
Many of these concerns are not addressed in conventional
100
pure steam measurements. A rule of thumb for turbine
effect is, for every one percent inlet moisture, the turbine Steam quality can be a prime indication of steam
stage efficiency drops by effectively one percent purity wider a certain state. However, this equality
(Kearton). requires correctioii when conditions change. Friction-
loss, heat-loss, stagnation state, precipi~tion,
Steam Purity dissolved solids in the vapor phase are prime factors
that influence this relationship. These corrections
Steam purity is the solidddissolved solid content in a must be addressed if precision of state is desired.
sample of condensed steam. Thus a 500 ppm TDS would
constitute steam condensate contains 500 part by weight
solids contamination per million total parts. Steam purity Representative Sampling
can be determined by evaporation, conductivity, flame
photometry or by trace chemistry means. In geothermal Representative steam sampling is not as simple as
service, trace chemistry is used to detect dissolved solids drawing a slipstream from a side port and having it
in small amounts in the condensate. analyzed. For even if the sample was analyzed
accurately, the results would be highly in error
Conductivity measurements are highly affected by because a representative sample was not taken.
dissolved gasses ("3 C02 HzS, etc.) in the liquid
fraction and as such is'not <sed for low TDS analysis. Unlike blended gas or liquid sampling that can be
The inajor impurity concerns are for silica, iron, representatively taken off side ports, multi-phase
chlorides, and arsenic. These elements are prime sampling of liquids, solids, gasses and steam is not
indicators of scaling, corrosion and SCC potential. Tracer evenly ~ s ~ b ~ withiii
t e d a conduit. They do not
ions for sodium or chlorides can provide a general ratio of behave as a homogeneous mixture. An understanding
the overall total dissolved solids (TDS) entering the of the flow regimes, thermodynamics and sampling
system once a baseline is established. probe designs are important for acquiring a
representative sample for analysis.
Typical fuel fired plants keep the silica content of the
steam below 25 ppb (Babcock & Wilcox). In geothermal Flow Regime
applications, the limit on silica concentration is restricted
to 1.0 ppm. Scale build-up can constitute reduced nozzle E a mixture of liquid and vapor was homogeneously
efficiency approximately 0.5% per 1/1000 inch of build- blended, with the liquid dispersed in sub micron
up (Kearton). particles, representative sampling would be simple.
~ ~ o ~ ~ a tthis e lisynot
, the case in steam quality
and purity sampling where at low rates the denser
liquid fraction wants to settle out on the bottom of
-
the pipe (Stratified Figure 1); at higher rates the
liquid concentrates onto the side of the pipe wall

576
 Jung
(Annular - Figure 2 ); and at even higher rates the
liquid droplets are sheared into small droplets and
distributed throughout the conduit ( Annular-Mist /
-
Disperse Figure 3 ). It is witlk this Amular Mist /
Disperse flow reghe that representative sampling
via tlie use of insertion sample probes can be tlie
most effective ( Flow Regime Map - Figure 4).
Fig. 2 Annular

Steam
R b b d
0
. ...
' . .
. e
0 .
.
*.*
.
. . . . ..
. . .. . . . .. ..
.

* =
. *
* *. * * * 0 .
- 0
e

.
.

0
* * *
.

Fig. 1 Stratified . . *. .::*

0 . .

*.
0 .

0
**
. . . *

Fig. 3 Mist/Disperse

Fig. 4 Baker Map

577
Jung
There are many obstacles that can alter the flow volume, traveling at a fixed rate, with a sample port
regime affecting sampling accuracy. These might iitercepting the path of all mass equal to this port
include: elbows, tees, valves, flow meters, size, that is isokinetic sampling ( Isokinetic Sampling
separators, thennowells, upflow, downflow, incline - Figure 5). If the sample rate is too fast, the port will
pipe positions,. etc. Dowfflow sampling is perferred suck more steam from the surrounding area diluting
if possible, followed by upflow, with horizontal lines the sample ( Fast Sampling - Figure 6). If the
being the more difficult to accurately sample. As a sampling rate is too slow, the heavier liquid mass will
general rule of thumb, a lower limit for sampling in penetrate the port concentrating the sample (Slow
horizontal lines is somewhere between 60% to 100% Sampling - Figure 7). And if the sampling rate is very
witli vertical mounted probes top and bottom slow, a stagnation boundary layer can develop in
preferred (Jug, 12). "lie lower tlie line pressure, the fiont of tlie probe deflecting a portion of the liquid
higher the required velocity to obtain mist flow. The fraction, causing a sample dilution (Stagnation Effect
higher tlie vapor velocity, the better the liquidhapor - Figure 8)(Jung, 14). The isokhietic sampling rate is
dispersion and tlie ability to sample representatively. determiled by the following relationship:

f = (&A) F (3)
Steam Sampling Nozzles
where:
There are two basic types of steam sampling probes:
tlie single point transversing, and the multiport f = sample flow rate.
stationary type. The transversing type can be more
accurate in undeveloped or complex profiles. Here F = totalflow rate in pipe.
the probe transverse across tlie pipe, much like a
pitot tube, sampling hi equal area segments. The d A = ratio of total port area to pipe area.
greater the number of samples, the better the
theoretical accuracy. In calorimeter use, a variable Care must be taken to insure equal sampling fiom all
extension tube is a source for heat-loss which is ports by internal design.
difficult to calibrate. The transversing technique is
best suited for chemical grab sample analysis. Its use
has an important feature in that wall fluid can be
monitored outlining impurity gradients.

The multiport sample probe is an impurity averaging Vapor

-----4
control
statioiiary design. Here a probe inserted into the
pipeline samples uniformly at equal area segments of
the pipe. It is important for structural stability that tlie
natural frequeiicy of tlie probe does not match the
Fig. 5 Isokinetic Sampling
forcing frequency (probe shedding vortex) of tlie
(Sample Velocity=Line Velocity)
steam flow or failure will occur. Some multiport
steam sampler probes are designed to operate
between the 1st and 2nd hannoiuc to reduce the
probe diameter aud interfering stagnation boundary
layer that can effect sampling accuracy (Jung, 14).

Isokinetic Sampling

Isohietic sampling means tlie flow velocity entering Fig. 6 Sampling Too Fast
each sample port must match die flow velocity w i t h (Dilute Sample)
the pipeline. This technique is iinportant in acquiring
a representative sample. If we consider a control

578
 Jung
where:
I I
x = initial steam quality,percent.

h2 = enthalpy of super-heated steam at

calorimeter (b tu/lb).
Fig. 7 Sampling Too Slow hg2 = enthalpy of saturated steam
(Concentrated Sample) @ calorimeter pressure.

hfl = enthalpy of saturated liquid in mixture

prior to throttling.

hf-1 = enthalpy of vaporization

corresponding to pressure or
temperature of steam entering
calorimeter.
Fig. 0 Stagnation Effect tc = superheated temperature in calorimeter
(Very Slow Sampling- Dilute)
t2 = saturation temperature ("F)
@ calorimeterpressure
Throttling Calorimeter
Cp = specific heat of saturated steam. Assume
The basic throttling calorimeter has been in use for 0.463for approximate atmospheric
over 100 years. It is a rather simple and old pressure in calorimeter.
teclmology used for tlie determination of steam
quality. The process functions by expanding steam
adiabatically across a nozzle. The initial and final Ai alternative way to determine steam quality is by
enthalpy will remain the same assuming there is no means of the Mollier Diagram ( Figure 9). Here point
heat-loss and kinetic energy changes. When wet A corresponds to the calorimeter superheat
saturated steam is throttled across a small colitrol telnPerature at SahlratiOn Pressure; POint

orifice, the liquid hction is atomized and vaporized correspond to a straight line to the initial line
into super-heat with the pressure drop from high- pressure. Tlis technique is a quick way to determine
pressure to low-pressure. By measuring the amount steam quality for moisture levels greater than 0.25%.
of super-heat existing within the calorimeter, the
amount of moisture can be determined. Design Styles

The basic relationship can be demonstrated by the Tliere are! many different calorimeter types such as
following equation: throttling, separating, throttlinglseparating and
separatinglthrottling. Our discussion will focus on the
throttling of wlich there are many designs. Tlie two
units that will be addressed will be the Ellison U Path
(Figure 10) and the ASME Pipe Fitting (Figure 11).
The U Path design is compact and incorporates a
or a more practical version: tliermo-barrier using the discharge steam and a quick
orifice change-out mechanism. The ASME is very
simple with several variations and can be
inexpensively fabricated with mostly off-the-shelf-
fittings. Regardless of the calorimeter used,

570
Jung
calibrating to mitigate various interferences that will
S e c t the steam quality determination is critical if Storm Pipe Nipplo
high precision is desired. A precision, calibrated
system can measure steam quality to 99.95%
(Jung, 14).

Heat Loss

Since tlie determination of steam quality is a function

of temperature, any heat transfer to the outside
control volume constitutes an error. Cold, wind and
rain all add to increase error. Insulation is critical on
tlie pipe, sampling appaatus and calorimeter. Even
witli insulation, the small sunout of heat loss is not
negligible for precision measurements above 99.5%.
Stationary and online calibration techuiques can
mitigate this effect.
I
Outlot Drain

FIGURE 10. Ellison U Path

IXHAUST

FIGURE 9

FIGURE 11. A.S.M.E. Pipe Flttlng

580
 Jung
Joule-Tbomson Effects Stagnation/Static State

Although some gasses will increase in temperature A coinmon misunderstanding in flow kinetics is the
when throttled, most gasses will experience a relationship between static and stagnation state. As
substantial decrease in temperature. The change in at! example, a dry saturated steam iri a stagnation
temperature over the change in pressure is known as (low velocity) state enters a nozzle. The steam
the Joule-Thomson Coefficient. That is: accelerating in a nozzle will not only experience a
drop in static pressure but will become wet ( i.e.,
Joule-ThomsonCoefficient = dT/dp (7) Wilson Line) ahd not super-heated (Moore, 10). This
is because some internal energy must be converted
Different gasses and vapors have different into kinetic energy and as a result, condensation or
coefficients when throttled. A mixture of gasses will supersaturation-exists.
not behave like a pure vapor altering the final
temperature. Geothermal steam contains gasses that
when throttled do not follow the same temperature / In geothermal systems because of seeding effects
pressure curve as pure steam. The greater the amount from impurities, the Wilson Line might not be
of gasses entrained in the steam, the more effect this reached before condensation occurs. If a sample is
will have on the throttling calorimeter readings. This theoretically taken in this high velocity wet zone, the
effect can be mitigated by calibratiou. results could be tainted as dry steam instead of wet.
The throttling calorimeter is basically a stagnation
Boiling Point Rise device. As a rule of thumb, if sampling is taken at
velocities greater than 150'/s, kinetic energy effects
This is generally a minor point except in areas where must be taken into consideration.
the brine has a high TDS such as in the Salton Sea
Geothermal Fields it1 California. Here the brine carry- Calorimeter Range
over entrained in the steam has a high dissolved solid
concentration that alters the boiling point of the fluid. The moisture range of the throttling calorimeter is
As such, vaporizing this fluid is more difficult, and restricted to its ability to produce a measurable
correcting for the solid content and scaling potential super-lieat (SH). This limit is approximately 3.15%
of the fluid presents other problems. @ 50 psig, 4.3% @ 100 psig, 5.6% @ 200 psig and
7% @ 400 psig. As a general rule of thumb, a
Gradients &num of 6°F SH to 10°F SH is required to
adequately vaporize all incoming moisture. This hnit
can be confinned visually by observing if the
In undersaturated liquids and superheated vapor
discharge calorimeter is wet or entrained droplets
temperature gradients are common. That is to say, if
with the steam are being emitted. If a higher moisture
one were to transverse a conduit or piece of
content is encountered (you could have serious
equipment, such as a calorimeter, there will exist a
temperature difference where a homogeneous problems), a separating calorimeter catl extend this
range.
average temperature might be desired. This effect can
be corrected by way of calibration.
Instrumentation
Barometric Pressure
Conventional throttling calorimeters depend on
As the barometric pressure changes, the temperature pressure and temperature measurements. Pressure
reading of the calorimeter can vary according to the gauge and transmitter accuracy are affected by
saturation temperature / pressure curve. Since stem temperature (Benedict). For every 10°F change over
quality calculations are based on the discharge calibration, the error can be affected up to 0.5%.
calorimeter pressure, this correction is important. Pressure gauges are generally psig and not psia
Online calibration can correct for this effect. calibrated, this effects accuracy with changing
atmospheric conditions. Since geothermal systems

58 1
Jung
incur variations in pressure, these variations must be For high precision measurements, a dual calorimeter
tracked to corresponding temperatures. system is required to compensate for fluctuatioiis in
line pressure, barometric pressure, ambient
High precision temperature measurements are temperature, gas variations, etc. One calorimeter will
preferred to pressure measurements for several provide the 100% steam quality reference, while the
reasons. Pressure fluctuations will be dampened by other tracks the moisture iU the steam. Everythmg
tlie mass of the temperature probe and correspond must be as d o r m as possible. The control orifice
more readily with other temperature elements. Partial must be calibrated to ensure isckinetic flow.
pressure of gasses will not have the same interfering Calculatioiis to size the orifice can be made using
effect on these elements. Having the same elements Napier's Formula:
of uniform calibration and response characteristics
enable more d o r m readings. Platinum RTD's of tlie
100 ohm, 3 or 4 wire shielded elements are preferred w = a1.97 WhenP2 is less than 0.58~1 (8)
for accuracy and stability. However, for differential 60
temperature measurement, thermocouple ( T - where:
Copper Coiistantm or J - Iron-Constantan) cut fiom
tlie same wire spool is more precise. Cold reference w = dry steam in pounds per second
junctions are required for thermocouples.
pi = upstreampressure psia
For temperature measurements, static and stagnation
conditions should be considered. It is important for P2 = downstreampressure psia
quick response to keep tlie mass of the temperature
element as small as practical. No themowells should a = smallest section of nozzle or orijice in
be used as the extra mass causes signdicant time square inches
delays for heat-up, slows response time, and acts as
an interfering heat sink. Element diameters of 1/4" to
1/8"stainless shielded stems are preferred. A mass flow calibration can be established by
flowing steam at choke flow through an orifice
Cali bration through a condensing coil (3/8" or 1/2") and
measuring the condensate / NCG's (Jung,l 1). As long
The calibration of a throttling calorimeter for as tlie down stream pressure of the orifice is
geothermal service requires establishing a control somewhere between 58% to 0% of the upstream line
baseline. Two measurements are required at any pressure, choke flow will be established and no
given temperature, pressure and flow rate: the additional flow will can pass though the orifice, even
temperature of the calorimeter (super-heat) with though a higher differential pressure is applied across
100% saturated steam entering; and the saturation the choke element.
temperature of the calorimeter at atmospheric
conditions. Tliis establishes the base reference point Application
where all saturated steam entering the calorimeter
must fall. Outside of this range and the moisture The throttling calorimeter is a valuable tool when
coiiteiit exceed tlie limitatioiis of the calorimeter or used properly. Its application complements precision
super-heat conditions exist. trace chemistry testing especially in areas where
moderate cost, fast response time, and continuous
A 100.00% steam quality reference source must be monitoring are desired. This tool can be applied to
established to calibrate the calorimeter. This requires monitor relatively dry production wells, separator
the same source steam that will be exposed to the performance, scrubber performance, steam wash and
calorimeter, at the same pressure, at the same steam turbine wash. The reaction time, for detecting a
flow rate (control orifice), an ultra high efficiency serious separator upset can be fast (in a matter of
polishing separator, with uniform dimensions and seconds), offering additional protection against
insulation on the external connections. turbine and facility damage fiom geothermal fluids.

582
 Jung
significantly improve accuracy in tlie
Summary determination of geothermal steam quality .
1. Poor geothermal steam quality & purity will 8: Throttling calorimeters incur errors fiom Joule-
cause premature equipment damage, reduce Thoinson effects, heat loss, boiling point rise,
power plant efficiency, and result in lost static / stagnation effects, atomization,
revenue. barometric pressure, gradients, etc., & sampling
inaccuracy.
2. Virtually all he1 fned steam power plants
continuously inonitor steam quality / purity or 9. Geotliemal steam quality can be measured to
fluid chemistry as a means to detect and prevent 99.95%with effective calibration and sampling.
scaling, corrosion, loss of efficiency and
premature damage to the steam turbine & 10. Throttling Calorimeters can provide an
components. inexpensive, fast response, continuous
monitoring instrument for the detection of
3. Representative steam quality and purity process upsets resulting 111 poor steam quality
sampling and analysis is intrinsically complex. and purity.
Isokinetic sampling, proper probe design, and
understanding flow regimes are critical.
Analytical techniques require special calibration
to compensate for interference of various References
contaminates and conditions unique to
geothemal applications. 1. ASME / ANSI PTC 19.11 Water & Steam in
the Power Cycle.
4. "hemodynamic & trace chemistry techniques
are two ways to analyze geothermal steam 2. ASME / ANSI PTC 19.3 Temperature
quality & purity. Both techniques can provide Measurement.
good accuracy and valuable information if
properly installed and calibrated. 3. ANSVASTM D1192 Equipment For Sampling
Water & Steam.
5 . Trace chemistry techniques, although inherently
of greater accuracy than thermodynamic 4. Benedict, R .P., Fundamentals of Temperature,
methods, if properly implemented, have long Pressure, and Flow Measurements, Second
time lags between sampling and analysis. The Edition, 1977, John Wiley & Sons.
accuracy is highly susceptible to sampling error
and contamhation if improperly implemented. 5. Moyer, J.A., Power Plant Testing, 4th Edition,
1934, McGraw Hill.
6. Geothermal facilities perform periodical grab
samples, with lag times between 6. Kearton, W.J.,Steam Turbine Theory and
sampling/analysis/resultsspanning hours to days. Practice, 7th Edition, 1958.
The results can give a false impression of high
steam quality and purity. Turbine scaling / 7. Babcock & Wilcox, Steam, 39th Edition, 1978.
erosion / corrosion is coinmon in geothermal.
service. 8. Baumeister & Mark, Standard Handbook For
Mechanical Engineers, 7th Edition, 1967,
7. Conventional calorimeter measurements in McGraw Hill.
geothermal applications can be highly erroneous.
Errors >> +/- 0.50% mass rates are common. 9. John, J.E.A., Gas Dynamics, 1969, Allyn &
This level of accuracy cannot protect the turbine Bacon.
and components fiom damage. Calibration
techniques described in this paper can

583
Jung
10. Moore, M.J., Two-Phase Steam Flow In
Turbines and Separators, 1976, Hemisphere
Publishing Corp.

11. Jug, D.B., Steam Quality & Purity Test Loop

Calibratiou & Apparatus, 1 1- 15-80, Intend
Report.

12. Jung, D.B., Steam Wash Pilot Test, 2-25-81,

Internal Report.

13. Juig, D.B., Sample Probe Design & Analysis, 5-

7-81, Internal Report.

14. Juig, D.B., Continuous Monitoring of Steam

Quality & Purity - Steam Wash, 10-16-81,
Internal Report.

15. Juig, D.B., BPP Turbine Scaling - Steam

Purification, 10-13-94, Internal Report.

584