Part 21 Report - 1998-080
ACCESSION #: 9801130266
ABB
January 6, 1998
LD-98-001
Document Control Desk
U.S. Nuclear Regulatory Commission
Washington, DC 20555
Subject: Infobulletin 97-08, "Information Update on the Use of the
DELTSTRAT Code for RCS Flow Rate Determination,"
Gentlemen:
As described more fully below, ABB Combustion Engineering (ABB-CE) has
issued the subject Infobulletin (attached) to its utility customers, and
has recommended that each utility using the DELTSTRAT code evaluate the
matters addressed in the Infobulletin for reportability under 10 CFR 21.
This report is provided to NRC for informational purposes.
The DELTSTRAT code is used at several ABB-CE plants (St. Lucie Units 1 &
2, Waterford Unit 3, San Onofre Nuclear Generating Station Units 2 & 3,
and Palisades) to determine RCS flow rate from data taken with a
calorimetric flow measurement technique. The code corrects for
temperature stratification effects in the RCS hot legs. A hot leg
temperature correction is calculated by the code, and is then applied to
the measured hot leg RTD temperatures to determine hot leg bulk coolant
temperature. The hot leg bulk coolant temperature, along with other
inputs, is then used to calculate the RCS flow rate.
In several instances, the code overestimated the hot leg temperature
correction and produced a calculated flow rate which was substantially
larger than expected. A review of test data and the underlying
assumptions for the DELTSTRAT code indicated that improvements in the
treatment of certain code inputs could be made which would produce a
calculated flow rate that would be closer to the expected value.
Infobulletin 97-08 discusses ABB-CE's findings and provides
recommendations to DELTSTRAT users for assuring that the code-calculated
RCS flow rates are reasonable and accurate.
ABB Combustion Engineering Nuclear Operations
Combustion Engineering, P.O. Box 500 Telephone (203) 688-1911
Inc. 1000 Prospect Hill Road Fax (203) 285-9512
Windsor, Connecticut
06095-0500
LD-98-001
Page 2
ABB-CE has evaluated the matters in the infobulletin with respect to 10
CFR 21 reportability, and has concluded that insufficient information is
available about how individual utilities use the DELTSTRAT code to
determine whether the issue is reportable under 10 CFR 21. Therefore, in
addition to the recommendation in the Infobulletin, ABB-CE has
recommended that each utility using DELTSTRAT evaluate whether the issues
addressed in Infobulletin are reportable under 10 CFR 21 in light of how
the individual utility uses the code.
Please do not hesitate to call me or George Hess at (860) 285-8405 if
there are questions concerning this matter.
Very truly yours,
COMBUSTION ENGINEERING, Inc.
Ian C. Richard, Director
Operations Licensing
Attachment: As stated
xc: S. Magruder (NRC)
ABB
Combustion Engineering Infobulletin No. 97-08
Dec. 22, 1997
Information Update on the Use of the DELTSTRAT Code for RCS Flow Rate
Determination
Summary: The DELTSTRAT code is used at several ABB-CE plants (St. Lucie
1&2, Waterford 3, SONGS 28,3, and Palisades) to determine RCS flow rate
from data taken with a calorimetric flow measurement technique. The code
corrects for temperature stratification effects in the RCS hot legs. A
hot leg temperature correction is calculated by the code, and is then
applied to the measured hot leg RTD temperatures in order to determine
hot leg bulk coolant temperature. The hot leg bulk coolant temperature,
along with other inputs, is then used to calculate the RCS flow rate.
In several applications, for Cycles 9 through 11 at Calvert Cliffs 2, the
code overestimated the hot leg temperature correction and produced a
calculated flow rate which was substantially larger than expected. A
similar application for Calvert Cliffs Unit 1, Cycle 12 produced an RCS
calculated flow rate which appeared to be more reasonable. The
overestimation of the RCS flow rate for Unit 2 appears to be related to
several factors: 1) unusual temperature trends in some of the hot leg RTD
measured temperatures, 2) the impact of a low leakage fuel management
scheme with a very peaked core radial power distribution, and 3)
isothermal biases for the RTD temperatures.
A review of the Calvert Cliffs test data and the underlying assumptions
for the DELTSTRAT code indicated that improvements in the treatment of
certain code inputs could be made which would produce a calculated flow
rate for Unit 2 that would be closer to the expected value.
This bulletin discusses the findings of the Calvert Cliffs review and
provides recommendations to the users for assuring that the DELTSTRAT
code calculated RCS flow rates are reasonable and accurate. A listing of
all assumptions inherent in the DELTSTRAT code is included here as
Attachment #1.
Discussion: There are several assumptions inherent in the formulation of
the equations and input for the DELTSTRAT code. One of the assumptions
is that the flow mixing factors, which relate how the flows exiting from
the core mix with adjoining flows as they proceed into the outlet nozzles
and down the hot legs, are invariant with time, power distributions, and
relative position of the reactor vessel internals. In the case of
Calvert Cliffs 2, for Cycles 9 through 11, the trend of measured hot leg
temperatures for channels A and C was unusual. The hot leg temperatures
for channels A and C indicated a constant core AT for these cycles even
though the expected trend, due to the low leakage fuel management, was an
increasing core AT. Channel B and D hot leg temperatures did show the
expected increasing trends with each cycle.
A second assumption used is that the normalized core radial power
distribution is an accurate proxy for describing the normalized fuel
assembly coolant temperature rise distribution (that is: the ratio of
individual assembly to core average coolant temperature rise). This
assumption works well for the traditional out-in fuel management schemes
used in the earlier fuel cycles for the ABB-CE plants. However, with the
more non-uniform radial power distributions inherent in some of today's
low leakage fuel management schemes, a better approach for defining the
fuel assembly coolant temperature rise is to use an open-core thermal-
hydraulic code, such as the ABB-CE TORC code, along with the actual core
radial power distribution. The open-core code will determine the open-
core effects of mass, momentum, and energy transfers between adjoining
fuel assemblies on the coolant temperature rise in each fuel assembly.
The assembly coolant temperature rise data can be normalized in the form
of assembly to core average temperature rise factors. The normalized
assembly coolant temperature rise factors can then be used as input to
the DELTSTRAT code (in place of the core radial power distribution).
_________________________
The information contained in this Infobulletin is provided by ABB-CE as a
service to your organization. Since operation of your plant is
completely within your control and responsibility, and involves many
factors not within ABB-CE's knowledge, this information may be utilized
only with the understanding that ABB-CE makes no warranties or
representations, express or implied, including warranties of fitness for
a particular purpose or merchantability, with respect to the accuracy,
completeness or usefulness of the information contained. ABB-CE
disclaims, and you assume, all liability, in negligence or otherwise, as
a result of your use of this information.
ABB-CE Infobulletin No. 97-08
Page 2 of 4 Dec. 22, 1997
A third assumption relating to the code inputs involves the measured cold
and hot leg coolant temperatures. The measured coolant temperatures are
assumed to have any inherent isothermal biases eliminated before being
input to the DELTSTRAT code. Removing biases will provide a more
accurate set of input to the code and a more consistent picture of the
temperature distribution in the RCS piping.
In the Calvert Cliffs Unit 2 situation, it was found that if the Channel
A and C temperature data were eliminated, and if open-core exit coolant
temperatures were used as input in place of the traditional core radial
power factors, the resulting calculated RCS flow rate would be much
closer to what was expected. The reasoning for eliminating the Channel A
and C data is as follows:
1. The DELTSTRAT code does not have the capability for predicting
unusual hot leg temperature trends which may be due to special,
localized effects as seen in the case of Calvert Cliffs Unit 2.
2. It is suspected that the measured hot leg temperature trend at
Calvert Cliffs Unit 2 is real, but local in nature and may not
be representative of the overall temperature distribution in
the hot leg.
3. By using only Channels B and D hot leg temperatures, a more
traditional temperature distribution was input to DELTSTRAT and
a lower calculated RCS flow rate was calculated.
Recommendations:
ABB-CE recommends that the users initiate the following actions in order
to assess if there are any current operability problems in determining
RCS flow rate in their plant(s) when using the DELTSTRAT code with
currently used procedures:
1. Review and confirm that the long-term trends (starting from BOC
1) in measured RCS flow rates, as determined with DELTSTRAT,
show consistency with any changes in plant hardware on the
primary side (i.e., with added SG tube plugging, RCP change-
outs, etc.), within the accuracy of the measurement method.
2. Compare current measured RCS flow rates with beginning-of-cycle
1 (BOC 1) values. If today's values are larger than the BOC 1
value by an amount which is greater than the measurement
uncertainty, without any corresponding justification from
hardware changes to the primary system, then consideration
should be given to reducing the RCS flow rate to the BOC 1
value. In the case of the digital plants, the flow rate in the
Core Operating Limit Supervisory System (COLSS) and Core
Protection Calculators (CPC's) would be reduced. For the
analog plants, the flow rate measured at the beginning of the
current cycle would be reduced.
If recommendation 1 or 2 fails, then proceed with recommendation 3 or 4.
3. Substitute normalized fuel assembly coolant temperature rise
fators, as determined by an open-core thermal-hydraulic
computer code, in place of a core radial power distribution
when preparing input to the DELSTRAT code, or else:
4. Determine a systematic flow bias that will be applied to the
flow rate calculated by DELSTRAT with core radial power
distribution as input. The flow bias will compensate for the
fact that the fuel assembly coolant temperature rise factors
are not being used as input.
_________________________
The information contained in this Infobulletin is provided by ABB-CE as a
service to your organization. Since operation of your plant is
completely within your control and responsibility, and involves many
factors not within ABB-CE's knowledge, this information may be utilized
only with the understanding that ABB-CE makes no warranties or
representations, express or implied, including warranties of fitness for
a particular purpose or merchantability, with respect to the accuracy,
completeness or usefulness of the information contained. ABB-CE
disclaims, and you assume, all liability, in negligence or otherwise, as
a result of your use of this information.
ABB-CE Infobulletin No. 97-08
Page 3 of 4 Dec. 22, 1997
It is recommended that for future cycles isothermal RTD temperature
biases be measured at the start of each new cycle for use in correcting
measured RTD temperatures prior to being input to DELTSTRAT. This step
is intended to improve accuracy, but is not required to support current
operability of the plants.
Applicability: All plants for which the ABB-CE DELTSTRAT code is used to
determine RCS flow rate.
Technical Contacts: R. P. Letendre (860) 285-2878
P. F. Joffre (860) 687-8062
Attachment #1 to Infobulletin No. 97-08
Assumptions used in the DELSTRAT Methodology
1. Eight hot leg and eight cold leg RTD's (from the safety channels)
provide input temperature values.
2. The average of the cold leg RTD readings in each RCS loop is equal
to the bulk cold leg temperature in that loop.
3. Hot leg and cold leg isothermal temperature biases are determined
and applied to adjust the measured RTD temperatures (individually or
collectively).
This is an assumption which the Infobulletin states is important in
obtaining unbiased temperatures for input to DELTSTRAT.
4. The flow mixing data from the ABB-CE Palisades flow model tests
apply to all ABB CE 2-loop reactor designs. The flow mixing factors
are invariant with time, core radial power distribution, coolant
temperatures, and any unusual localized flow conditions in the hot
legs.
This is an assumption which appears to be challenged by the Channel
A and C hot leg temperature data trends at Calvert Cliffs Unit 2
during Cycles 9 to 11.
5. The core radial power distribution is a representative proxy for the
normalized distribution of fuel assembly coolant temperature rise
factors.
This is an assumption which the Infobulletin states can be improved
upon by using a calculated normalized radial distribution of fuel
assembly coolant temperature rise factors as input to DELTSTRAT in
place of the radial power distribution.
6. Flow mixing in the reactor vessel upper plenum between the two loops
is negligible.
_________________________
The information contained in this Infobulletin is provided by ABB-CE as a
service to your organization. Since operation of your plant is
completely within your control and responsibility, and involves many
factors not within ABB-CE's knowledge, this information may be utilized
only with the understanding that ABB-CE makes no warranties or
representations, express or implied, including warranties of fitness for
a particular purpose or merchantability, with respect to the accuracy,
completeness or usefulness of the information contained. ABB-CE
disclaims, and you assume, all liability, in negligence or otherwise, as
a result of your use of this information.
ABB-CE Infobulletin No. 97-08
Page 4 of 4 Dec. 22, 1997
7. Complete mixing of the hot leg coolant temperatures occurs at a
distance of 18 hot leg pipe diameters downstream of the reactor
vessel outlet nozzle exit.
8. The circumferential distribution of hot leg coolant temperature can
be described by a cosine variation.
9. The flow in the hot legs may rotate in a solid-body manner as it
passes down the length of the hot legs.
10. The velocity distribution in the cross-section of the hot legs for
Plant X is the same as existed in the Palisades reactor flow model
tests, from which the flow mixing factors were determined.
List of recently-issued Infobulletins:
Number Date Title
97-07, Rev 01 12/31/97 Tech Spec on Azimuthal Tilt in Analog Plants
97-07 10/28/97 Tech Spec on Azimuthal Tilt in Analog Plants
97-06 10/21/97 Core Snubbers / Blocks
97-05 10/21/97 Core Ledge Loads
97-04 7/11/97 Potential Error in the Energy Redistribution
Factor Used in LOCA Analysis
97-03 6/5/97 Use of Non-Safety (non-Q) Components in Reactor
Protection System
97-02 5/23/97 Spurious Recirculation Actuation Signal
97-01, Supl 1 4/9/97 CESEC Decay Heat Model
97-01 4/3/97 CESEC Decay Heat Model
96-02 4/24/96 Steam Generator Sleeve-to-Tube Weld Indications
_________________________
The information contained in this Infobulletin is provided by ABB-CE as a
service to your organization. Since operation of your plant is
completely within your control and responsibility, and involves many
factors not within ABB-CE's knowledge, this information may be utilized
only with the understanding that ABB-CE makes no warranties or
representations, express or implied, including warranties of fitness for
a particular purpose or merchantability, with respect to the accuracy,
completeness or usefulness of the information contained. ABB-CE
disclaims, and you assume, all liability, in negligence or otherwise, as
a result of your use of this information.
*** END OF DOCUMENT ***
Page Last Reviewed/Updated Tuesday, March 09, 2021