Proceedings

Proceedings ofof
ASME TURBO EXPO 2002 – Land, SeaEXPO
ASME TURBO 2002
& Air 2002
June 3-6,
June2002, Amsterdam,
3-6, 2002, Amsterdam,The
The Netherlands
Netherlands

GT-2002-30482

THE MODEL STEAM TURBINE: MORE DETAILS ABOUT THE HEAT RECOVERY
STEAM GENERATOR EVALUATING METHOD IN A COMBINED CYCLE PLANT

Dr. Dietmar Schmidt Michail Arnold

Siemens PG, Erlangen, Germany Siemens PG, Erlangen, Germany

ABSTRACT Another positive aspect of this calculation method is that it

Turnkey and thermal island supply scopes present turbine evaluates the worth of individual HRSG steam outputs to the
suppliers with a perfect way to sell their rotating products. The cycle, and thus is a fair “thermodynamic compensation”
popularity of these plant configurations, along with the recent mechanism.
availability of more holistic test codes, has led to the need for
an accurate and reasonable method of determining the thermal NOMENCLATURE
performance of the externally-purchased HRSG component. DRE Data Reduction Equation
To assess a multiple pressure HRSG, it is advantageous and f loss Heat Loss
convenient to have one single criterion for the evaluation of
performance, especially when this criterion provides for the m Mass Flow
compensation of the different outlet energy streams. The so- OEM Original Equipment Manufacturer
called Model Steam Turbine method of HRSG evaluation was P Power
developed for these reasons. The result of the calculation, a p Pressure
lone performance criterion, is the shaft power of the fictitious Q Heat Flow
Model Steam Turbine. T Temperature
v Specific Volume
INTRODUCTION UMF Uncertainty Magnification Factors
The model steam turbine method of HRSG evaluation was UPC Uncertainty Percentage Contributions
developed specifically to test steam generators in combined
ηi, ST Steam Turbine Expansion Efficiency
cycle plants. Being an integral part of the combined cycle, the
HRSG must perform properly. As it is supplied by the OEM to
the owner of the combined cycle guarantees, the performance of Indices
this component must be demonstrated individually to this Corrected Corrected Value
owner. Design Design Value
The current standard of steam generator testing DIN 1942 CRH Cold Reheat
[3] is not clearly adopted for HRSGs. ASME PTC 4.4-1981 GT Gas Turbine
provides different methods of HRSG evaluation. However, none HRH Hot Reheat
of these standards provide a satisfactory algorithm for HRSG HRSG Heat Recovery Steam Generator
evaluation which shows clearly the impact of the HRSG steam HP High Pressure
output on the combined cycle plant performance. In Inlet
Using the presented test method, a straightforward Isentropic Isentropic Expansion
evaluation yields the HRSG performance at guarantee LP Low Pressure
conditions. As this performance value is given in units of power Measured Measured Value
from the steam turbine, it directly addresses what is important to MST Model Steam Turbine
the combined cycle thermal performance guarantees. Out Outlet

1 Copyright © 2002 by ASME

1 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Figure 1: Measurement Point Diagram for a typical 3-pressure HRSG with reheater and fuel gas heating

Assumptions. Even with all of these measurement points,

CALCULATING MODEL STEAM TURBINE POWER we will still need to make some assumptions about this steam
In order to arrive at a MST power for a HRSG, several generator. The heat losses due to radiation, convection and
calculation steps are required. We will identify them as: conduction will have to be approximated from either a detailed
necessary measurements and assumptions, the gas turbine calculation based on steam generator surface area and
exhaust mass flow calculation, the exhaust gas composition insulation, or the guidelines of a standard such as DIN 1943 or
calculation, the corrections, and the MST power calculation. ASME PTC 4.4.
In the event that cold reheat flow is not directly measured,
Measurements. Proper HRSG testing requires a lot of it is necessary to find a reasonable value by either assuming a
measurements. The exact number varies, but, as Figure 1 shows, constant reheat ratio (the amount of HP steam which survives to
all energy streams entering and leaving the steam generator become cold reheat steam), or by estimating the actual leakage
need to be considered. Figure 1 contains over 40 measurement via a comparison of design and test volumetric flow rates and
points; a number which could easily translate into 80 primary pressures.
measurement devices once the exhaust gas inlet and outlet If proper care is taken to ensure that drum levels are
temperature grids are considered. constant during testing, then one can get away with only
Refer to Figure 1 for a Measurement Point Diagram of a measuring either the feedwater flows or the steam flows,
typical 3-pressure HRSG with reheater and fuel gas heating whichever has a better accuracy.

2 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 If no accurate measurement of the condensate flow is piping pressure drops are properly accounted for when
available, feed water mass balance calculations can be used to generating this set of corrections.
calculate this parameter. If the test goal includes providing information about the
A small factor in the exhaust gas flow calculation (by performance of individual pressure levels in addition to the
HRSG energy balance method) is that of re-circulation pump determination of MST power, then it is advisable to structure
power consumption, which can be assessed as the rated power the corrections so that each HRSG output to the ST is corrected
consumption. individually. In this way, more analysis about the internal
performance of the steam generator is possible. These values
Exhaust gas mass flow calculation. One of the challenges are then inserted into the MST algorithm for the demonstration
to HRSG testing is the accurate determination of exhaust gas of MST power. As this method provides more detailed
mass flow. The MST method of HRSG evaluation is flexible in information about the HRSG, it is preferred by the authors.
this respect, being conducive to gas turbine exhaust flow Keeping in mind that plant configuration and scope-of-
calculation by a variety of methods: GT or HRSG energy supply vary, a sample list of applicable correction parameters
balance, O2 measurement, etc. Using the HRSG energy balance for this test is provided:
method, which we prefer for reasons to be discussed later, 1. Exhaust gas mass flow
yields the following equation: 2. Exhaust gas temperature
3. Exhaust gas enthalpy or cp (at design gas temperature)
Q HRSG − PPumps 4. HP steam pressure
m GT= (1) 5. Cold reheat steam mass flow
æ f ö 6. Cold reheat steam temperature
(hHRSG, in − hHRSG, out ) ⋅ çç1 − loss ÷÷ 7. Hot reheat steam pressure
è 100% ø 8. LP steam pressure
9. Fuel gas heater heat exchange
where f loss is the heat losses due to radiation, convection, 10. Temperature difference between condensate preheater
conduction. outlet and HP feedwater inlet (may be negligible)
It is here noted that the measurements necessary for this 11. Temperature difference between condensate preheater
calculation are almost identical to those required for measured outlet and IP feedwater inlet (may be negligible)
steam variables and corrections. 12. Condensate inlet temperature

Exhaust gas composition calculation. The determination MST power calculation. To compute the shaft power of
of exhaust gas constituents is required for both the exhaust gas the Model Steam Turbine as a performance guarantee of the
mass flow calculation and the performance correction for HRSG, the expansion efficiencies of the three pressure stages
exhaust gas composition (described as either cp, enthalpy, or from the guarantee heat flow diagram of the Combined Cycle
H2O content). DIN 1942 offers a reasonable method, called the Plant are used. This is a very reasonable assumption, as the
‘Ultimate Analysis’ method, which can determine constituents steam turbine efficiency is virtually constant for small changes
from measurements of stack O2, ambient air conditions, and the in inlet conditions.
ultimate analysis of the fuel. One advantage of this method is The corrected steam piping pressure drop between the
that it does not require the accurate measurement of fuel gas HRSG outlet and the MST inlet is calculated based on the ratio
flow to the gas turbine. of corrected and guaranteed volumetric flow rate.
The corrected HRSG outlet pressure is calculated with
Corrections. As with any performance test, corrections Stodala’s law using design and measured steam flows, design
must be applied to the measured performance to determine if and measured density, design piping pressure loss and design
the HRSG has achieved the guaranteed MST power. Depending turbine outlet pressure.
on the goal of the test, these corrections can either be applied The corrected steam turbine inlet pressure is then
directly to the MST power calculated from test measurements, calculated from the corrected HRSG outlet pressure and
or to each measured steam parameter which contributes to the corrected steam piping pressure drop values.
MST power.
If the aim of the test is simply to demonstrate MST power, Steam Piping Pressure Drop Calculation. For the
then much time and calculation effort can be saved by pressure drop calculations an adiabatic steam transport is
structuring the corrections so that they apply directly to the assumed in the steam pipes.
MST result. Instead of 6 complete sets of corrections
(HP/HRH/LP steam flow and temperature), only one is needed.
On a cautionary note, care must be taken to ensure that the

3 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 m i2, corrected vi , corrected
∆pi , corrected = ∆pi , design ⋅ 2 ⋅ (2) pi, ST in, corrected =
m i , design vi , design
m i, ST, corrected
⋅ (pi,2 ST in, design − pi,2 ST out, design ) (5)
where index i represents HP, IP / Hot RH and LP m i, ST, design
alternatively
∆pi , corrected measured and corrected pressure drop + pi,2 ST out, corrected
∆pi , design design pressure drop
The above formula is used to determine the turbine inlet
m i , design design steam mass flow pressures for the HRSG acceptance test model turbine power
output.
m i , corrected measured and corrected steam mass flow
Using the above mentioned formulas (4) – (5), the steam
vi, design specific volume = f ( pi, design ,Ti, design ) parameters used for the test model steam turbine power can be

f ( pi , design ,Ti , corrected )

calculated starting from the fixed end point of evaporation in
vi , corrected specific volume = the condenser.

So, the Model Steam Turbine inlet pressure for each Calculation of the MST Expansion Efficiency. The
pressure level is calculated as: compiled data in the corresponding heat flow diagram are used
for the calculation of the MST expansion efficiency.
pi, ST in, corrected = pi, ST in, measured − ∆p i, corrected (3)
hi, ST in, design − hi, ST out, design
ηi, ST = (6)
Calculation of Turbine Inlet Pressures. The measured hi, ST in , design − hi, ST out , isentropic
and corrected HRSG outlet mass flows have got an influence on
the pressure reduction of a steam turbine. The following
where index i represents HP, HRH and LP alternatively
Stodola-Equation describes the ratio between design mass flow
and pressure reduction to measured mass flow and pressure ηi, ST ST expansion efficiency
reduction as follows: hi, ST in, design ST design inlet enthalpy

pi,2 ST in, design − pi,2 ST out, design hi, ST out, design ST design outlet enthalpy
= hi, ST out, isentropic ST outlet enthalpy
m i, ST, design
(4) (h-s-diagram: isentropic expansion)
pi,2 ST in, corrected − pi,2 ST out, corrected As described above, with a constant expansion efficiency
m i, ST, corrected assumed, the Steam Turbine outlet enthalpy for the HRSG
acceptance test can be calculated with this formula:
where index i represents HP, IP / Hot RH and LP
alternatively
hi, ST out, corrected = hi, ST in, corrected −
pi,2 ST in, design Design Turbine Inlet Pressure (h i, ST in, corrected − hi, ST out, isentropic, corrected )⋅ ηi, ST
(7)

pi,2 ST out, design Design Turbine Outlet Pressure

where index i represents HP, HRH and LP alternatively
m i, ST, design Design Turbine Mass Flow hi, ST out, corrected ST outlet enthalpy (model turbine)
p 2
i, ST in, corrected Calculated Turbine Inlet Pressure hi, ST in, corrected ST inlet enthalpy
pi,2 ST out, corrected Calculated Turbine Outlet Pressure = f ( pi, ST in, calculated , Ti, ST in, corrected )
(start with design condenser pressure) hi, ST out, isentropic, corrected ST outlet enthalpy [kJ/kg]
m i, ST, design Measured Turbine Mass Flow (isentropic expansion, h-s-diagram)

As the characteristic of turbine pressure reduction builds up These calculations are carried out for each pressure stage.
from the outlet to the inlet the above formula leads to the The overall Steam Turbine Shaft Power, the criterion for the
following formula for the measured turbine inlet pressure: HRSG acceptance test, is determined by the following formula:

4 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 uncertainty, and can have an impact on the
Measured Values Xn Individual Uncertainty
structure of the corrections as well.
of Measured Values Xn
Thermodynamic Compensation. Implicit in
the MST calculation is thermodynamic
X1, X2, ..., Xi-1, Xi ± MUi, ...,Xn compensation between the energy streams leaving
MUi = Measurement Uncertainty of Xi the HRSG test boundary. The decrease in MST
n sets for n values output caused by the under-performance of, say,
HP steam flow could be offset by the added
output that an over-abundance of HRH steam
Heat Recovery Steam Generator
temperature yields. Better still is the fact that the
Energy Balance Combustion Calculation MST calculation automatically assesses the
proper “value” that each guaranteed steam
Exhaust Gas h or H2O-content parameter holds for the combined cycle; namely,
Mass Flow of Exhaust Gas the amount of steam turbine power it can
produce. A further extension of the
Correction Curves thermodynamic compensation principle is
Corrected possible in the case where multiple HRSGs are
feeding the same turbine. For this case, the
Model Steam Turbine control boundary could be drawn around both
steam generators, with their performance being
evaluated as if they were one steam generator.
Before this “super steam generator” test boundary
Test Power Output of could be applied; however, it would be necessary
Model Steam Turbine to determine that averaging of test boundary
Yi = f(X1, X2, ..., Xi-1, (Xi ± MUi), ..., Xn) conditions did not decrease the accuracy of test
n sets for n values results.
Test Power Output of
Overall Measurement Uncertainty Method of Uncertainty Calculation.
Model Steam Turbine
Y = f (X1, X2, ..., Xn) [Σ (Y-Yi) * 100 / Y)2 ]0,5 Figure 2 is used to outline a method for correctly
determining the uncertainty of the test result from
the measured variables per the given standard [5].
The following assumptions can be made:
1. MST power is the proper parameter for
evaluation; therefore, the MST power is
the result of interest.
Final Results 2. The random errors are negligible relative
to the systematic errors.
Figure 2: Scheme of the algorithm of the Measurement Uncertainty Analysis 3. There are no significant correlation
terms.
Based on the above assumptions, the calculation of the
( )
PMST = hHP, ST in, corrected − hHP, ST out, corrected ⋅ m HP, corrected overall test uncertainty can be made by
( )
+ hIP, ST in, corrected − hIP, ST out, corrected ⋅ m IP, corrected (8)
• using a central-difference numerical scheme to
calculate the partial derivative terms.
( )
+ hLP, ST in, corrected − hLP, ST out, corrected ⋅ m LP, corrected • Setting a step size to calculate the partial derivative
terms.
where
• Calculating Uncertainty Magnification Factors (UMF)
PMST Steam Turbine Shaft Power and Uncertainty Percentage Contributions (UPC) to
gain more insight into the results of this analysis.
MST AS A METHOD OF HRSG EVALUATION Each variable needs to be checked for its sensitivity to step
Testing according to the MST concept is different in many size. The uncertainty term can be calculated by multiplying the
ways. In addition to the automatic thermodynamic Partial derivative with the uncertainty estimate. To calculate the
compensation which occurs, this method will change the test total absolute uncertainty in the MST power, the root sum
square of all of the uncertainty terms must be build.

5 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 To calculate UMF the following formula is used: to the uncertainties of calculating steam parameters
individually.
X i ∂r This reduction in test uncertainty is realised when HRSG
UMFi = ⋅ (9) energy balance is used to calculate exhaust gas flow en route to
r ∂X i determining the MST power. The reason for this is a sort of
“self-correcting” mechanism in the program, with logic as
∂r follows:
where Xi is the measured variable, is the partial 1. HP mass flow is measured 1% higher than actual
∂X i 2. Exhaust gas mass flow is calculated 0.6% higher than
derivative for the measured variable, and r is the result (MST actual
power). 3. Higher exhaust gas mass flow yields in an higher
The UMF indicates the influence of the uncertainty in that expected HP mass flow, which corrects HP mass flow
variable on the uncertainty in the result. A value greater than 1 downwards.
indicates that the uncertainty of the variable is magnified as it 4. The correction factor for exhaust gas mass flow also
propagates through the data reduction equation (DRE) into the corrects the HRH and LP steam flows downward.
result. A value less than 1 indicates that the uncertainty is 5. Calculated MST power increases 0.18%.
diminished. The sign is unimportant since all terms are squared
for the uncertainty calculation. Note that these terms do not A detailed scheme of the algorithm of the Measurement
include uncertainty estimates. They are for the general case with Uncertainty Analysis is shown in Figure 2.
the specific DRE to be used.
To calculate UPC the following formula is used: Corrections. As mentioned before, the option to apply
corrections directly to the measured model steam turbine power
2
æ ∂r ö instead of to the individual steam parameters can reduce the
çç ÷÷ (U xi )2 amount of work involved with generating these corrections.
∂X
UPCi = è i ø 2 ⋅100 (10)
Ur Instrumentation. No additional measurements are
necessitated by the MST method of HRSG evaluation. The
decision of how to calculate exhaust gas flow, on the other
where is hand, leads to changes in the measurement point list. Fuel flow
∂r the partial derivative for the measured and electrical metering come with a GT energy balance, while
∂X i variable the HRSG balance method calls for an HRSG exhaust
the uncertainty estimate for the measured temperature grid. One possible advantage to the latter method is
U xi variable that it allows for HRSG testing to be conducted without the GT
Ur the uncertainty of the result (MST power). instrumentation which is often supplied by the GT OEM and
not the HRSG OEM.
The UPC for a given X i gives the percentage contribution
of the uncertainty in that variable to the squared uncertainty of Exhaust Gas Calculation. As mentioned before, the MST
the result. It includes the effect of both the UMF and the evaluation method can be realized with any type (or
magnitude of the uncertainty of the variable. All UPC values combination of types) of exhaust gas flow calculation. The
sum to 100 percent of the uncertainty of the result. authors feel that this flexibility is attractive when compared to
The UMF and UPC values should provide some insight to the limitations of present test codes.
the critical factors in accurately determining the MST power.
SOME NOTES ON CORRECTIONS
Test Uncertainty. Because it incorporates all energy The only way to gather meaningful data about the steam
streams simultaneously, an MST-based performance test can generator is to correct the measured performance to guarantee
achieve a lower test uncertainty than a test in which individual conditions. Whether this is done with thermal model calculation
steam parameters are evaluated. or by the application of correction curves / equations, it is
From a mathematical standpoint, the lower sensitivity necessary to ensure that all the applicable test boundary
coefficients which each steam parameter has in the MST power parameters are considered in the corrections. As this subject is
calculation will also serve to reduce test uncertainty. The not comprehensively discussed as specifically pertains to
sensitivity coefficient of the measured HP steam flow on the HRSGs in either the current ASME PTC 4.4 or DIN 1942 test
corrected HP steam flow is 1.0%/%. The sensitivity coefficient codes, a few concepts are included in this paper.
of the measured HP steam flow on the corrected MST power is
close to 0.18%/%. These small sensitivity coefficients tend to Hot reheat steam pressure. Generally, both hot and cold
yield an improved uncertainty for the MST output as compared reheat pressures can be used as a correction of all steam

6 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 parameters. The authors prefer the correction with the hot CONCLUSIONS
reheat steam pressure because the hot reheat steam pressure Using the model steam turbine to compare guarantee and
depends of the IP section of the Steam Turbine. The cold reheat tested HRSG performance is both relevant and fair. The result
pressure consists of the hot reheat pressure plus the reheater of interest, namely how much useful steam energy is brought to
section pressure loss. The cold reheat pressure affects the the steam turbine by the steam generator, is accurately evaluated
expansion of the HP steam turbine section. Because of this, the with this method.
cold reheat pressure can not be used as an independent variable Additionally, test uncertainty can be reduced if the exhaust
for the correction. gas is calculated by HRSG energy balance.
As this method provides for thermodynamic compensation,
Condensate pre-heater. One possible plant configuration it is fair to both the supplier and purchaser of the HRSG
is one in which the condensate flow leaves the HRSG test component.
boundary in-between the condensate pre-heater outlet and the
economizer inlets, as is the case when the feedwater pumps ACKNOWLEDGMENTS
and/or de-aerator(s) are external to the steam generator The authors would like to thank Susan T. Hudson,
supplier’s scope-of-supply. With such a configuration, it may be Mississippi State University for her expertise in Measurement
necessary to correct the measured plant performance if the Uncertainty Analysis.
relationship between the exiting and re-entering fluid streams is
other than design. In this case, the authors recommend a REFERENCES
correction which is based on the temperature difference [1] Dietmar Schmidt, Terrence Sullivan:
between the condensate pre-heater outlet stream and the The Model Steam Turbine: A method of evaluating a
economizer inlet stream temperatures. Such a correction would Heat Recovery Steam Generator in a Combined Cycle
be expected to have a low impact on the test result. Plant
Proceedings of ASME TURBO EXPO 2001, June 4-7,
Exhaust gas constituents. Several ways of correcting for 2001, New Orleans, Louisiana, 2001-GT-0501
exhaust gas composition can be used. The most comprehensive [2] DIN 1943
method, that of correcting for individual constituents, can be “Thermal Acceptance Tests of Steam Turbines”.
done easily in a thermal model but is impractical if individual English version.
corrections are used each constituent would need a separate Published 1975
correction. Instead, most HRSG suppliers generate corrections [3] DIN 1942
based on either exhaust gas water content, exhaust gas enthalpy, “Acceptance Testing of Steam Generators”.
or exhaust gas cp. Of these methods, the authors prefer to use English version.
the latter two. Correcting based on exhaust gas water content is Published 1994
close but not perfect; because it does not consider the minor [4] ANSI / ASME PTC 4.4-1981
differences in specific heat of the non-water constituents. When Gas Turbine Heat Recovery Steam Generators
correcting for either enthalpy or specific heat at constant Published 1981
pressure, all calculations should be based on the design [5] ASME PTC 19.1-1998
temperatures, not the measured temperatures during the test. Test Uncertainty – Instruments and Apparatus
This is done to avoid applying a double correction for the Published 1998
impact of HRSG inlet exhaust gas temperature.

7 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/02/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use