You are on page 1of 367

Method of Reducing Carry-Over and Reducing

Pressure Drop Through Steam Separators

NP-1607
Research Project S122-1

Final Report, November 1980

Prepared by

COMBUSTION ENGINEERING, INC.


C-E Power Systems
Chattanooga Nuclear Operations
Chattanooga, Tennessee 37402

Principal Investigators
W. R. Carson
H. K. Williams

Prepared for

Electric Power Research Institute


3412 Hillview Avenue
Palo Alto, California 94304

EPRI Project Manager


D. A. Steininger

Steam Generator Project Office


Nuclear Power Division

0**
DISCLAIMER

This report was prepared as an account of work sponsored by an


agency of the United States Government. Neither the United States
Government nor any agency thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal liability
or responsibility for the accuracy, completeness, or usefulness of any
information, apparatus, product, or process disclosed, or represents
that its use would not infringe privately owned rights. Reference
herein to any specific commercial product, process, or service by
trade name, trademark, manufacturer, or otherwise does not
necessarily constitute or imply its endorsement, recommendation, or
favoring by the United States Government or any agency thereof. The
views and opinions of authors expressed herein do not necessarily
state or reflect those of the United States Government or any agency
thereof.

D IS C L A IM E R

Portions of this document may be illegible in electronic image


products. Images are produced from the best available
original document.
EPRI authorizes the reproduction and distribution of all or any portion of this report and the preparation
of any derivative work based on this report, in each case on the condition that any such reproduction,
distribution, and preparation shall acknowledge this report and EPRI as the source.

NOTICE
This report was prepared by the organization(s) named below as an account of work sponsored by the
Electric Power Research Institute, Inc. (EPRI) and the Steam Generator Owners Group. Neither EPRI,
members of EPRI, the Steam Generator Owners Group, the organization(s) named below, nor any
person acting on their behalf: (a) makes any warranty or representation, express or implied, with respect to the
accuracy, completeness, or usefulness of the information contained in this report, or that the use of any
information, apparatus, method, or process disclosed in this report may not infringe privately owned rights; or
(b) assumes any liabilities with respect to the use of, or for damages resulting from the use of, any information,
apparatus, method, or process disclosed in this report.
Prepared by
Combustion Engineering, Inc.
Chattanooga, Tennessee
EPRI PERSPECTIVE

PROJECT DESCRIPTION

Performance of steam-water separators in steam generators is a subject of continuing


concern. Field modifications to increase steam generator circulation ratio, and
thereby decrease sludge deposition and the tendency for dryout and chemical concen­
tration, have increased the moisture loading on primary separators and sometimes
caused excessive moisture carry-over. There has also been interest in increasing
steam generator circulation ratio by reducing the resistance to flow in the primary
separators themselves. Finally, the scaling from small prototypes to the larger
separators in steam generators has sometimes been inappropriate, with the result
that the larger separators have failed to meet operational criteria. This final
report for RPS122-1 provides an overview of separator design. RPS172-1 and
RPS173-1, other Steam Generator Project Office projects, address separator analysis
and testing.

PROJECT OBJECTIVES

This report is a comprehensive compilation of the state of the art dealing with
steam-water primary separators. An extensive 1iterature review was developed, and
separator designs worthy of continued development were identified. Additionally, a
systematic test program for the qualification of these advanced separator designs
was defined.

PROJECT RESULTS

This report represents a formidable effort to document in one place the extensive
but diffuse information on primary steam-water separator development, testing, and
design. It can be concluded from this investigation that separator designs have
been developed in the past by trial-and-error and confirmatory testing. Little
analytic work in conjunction with experimental qualification testing has been
performed. To put the design of primary separators on a firm analytic foundation,
work should be directed toward developing analytic techniques that will accurately
describe the performance and hydraulics of primary separators. This report should
provide a foundation upon which such an advanced effort can build and should be of

iii
interest to steam generator designers and utility personnel responsible for steam
generator reliability.

D. A. Steininger, Project Manager


Nuclear Power Division

iv
ABSTRACT

Steam separation equipment affects both the operation and performance of a PWR
steam generator in terms of pressure drop, carryover, and carryunder. Carryover,
the moisture entrained in the exiting steam flow, results in decreased plant
efficiency and may lead to premature turbine blade failure. Carryunder, steam
entrained in the circulating water flow, has a potential for decreasing the
circulating water flow rate, and it may adversely affect the operating perfor­
mance of the steam generator. To help prevent chemical hideout in the tube
bundle, a high circulating water flow (i.e., circulation ratio) should be deve­
loped. The differential pressure for this circulating flow cannot be maintained
if excessive carryunder exists which cannot be condensed by the addition of sub­
cooled feedwater in the steam generator downcomer. Recognizing these principles
of operation, this project determined the benefits which might be recognized from
improved primary steam generator performance, identified three separator designs
worthy of future development, and developed a recommended test program for veri­
fication testing of separator design improvement. In general, however, the
studies revealed that the circulation ratio cannot be significantly improved by
decreasing the primary separator pressure drop. Backfit of presently used pri­
mary separators into field steam generators to obtain improved performance was
evaluated. An extensive literature search in the area of separator design was
also developed.
CONTENTS

Section pages
1 INTRODUCTION 1-1
2 SUMMARY 2-1
3 RESULTS 3-1
4 OPTIONALSTUDIES AND TESTS 4-1
APPENDIX A LITERATURE SURVEY A-l
APPENDIX B PRESENTLY USED SEPARATOR EVALUATION B-l
APPENDIX C DRYER DESIGN FEATURES AND PERFORMANCE C-l
APPENDIX D ALTERNATIVES TO TWO STAGE SEPARATION D-l
APPENDIX E DESIGN FEATURES E-l
APPENDIX F RECOMMENDED SEPARATOR CONCEPTS F-l
APPENDIX G PROMISING SEPARATOR CONCEPTS G-l
APPENDIX H RECOMMENDED TEST PROGRAM H-l
APPENDIX I SPACE/SIZE LIMITATION EVALUATION 1-1
APPENDIX J HYDRAULIC EVALUATION J-l

vii
Section 1

INTRODUCTION

The objectives of this program were:

1. Develop a thorough understanding of the parameters which affect


primary separator performance and the related affects of separator
performance and pressure loss on steam generator operation, espe­
cially on recirculation flow.

2. Determine the potential benefits to be derived from further primary


separator development.

3. Identify a 1imited number of promising improvements in steam sepa­


rator design which warrant testing for further refinement and per­
formance verification.

4. Recommend a test program for verification testing of identified


separator improvements.

To satisfy the four objectives, this program focused on determining the increase
in steam generator circulation flow rate resulting from decreasing the primary
separator pressure loss, determining the design features which would permit in­
creasing the steam loading capacity of primary separators while meeting accept­
able carryover and carryunder performance, and determining design improvements
which have the potential of reducing primary separator water carryover and carry
under.

The initial emphasis of this program was to determine whether a decreased separa
tor pressure loss could significantly increase circulation rate (CR), thereby
resulting in improved steam generator hydraulics. A hydraulic analysis of
various U-tube steam generators indicated that a significant gain in circulation
ratio could not be obtained by decreasing separator shock loss (pressure loss).
The main reason for this is that only approximately 20% of the pressure loss in
the circulation loop is in the primary separator.

In the latter half of the program, the emphasis shifted toward improving the
separator carryover performance and increasing the separator flow loading capa­
bility. Of those separators evaluated, those with axial vanes operated with

1-1
higher total loading (steam and water flow rates) for given values of carryover
and carryunder. The separators with the centrifugal vanes had low carryover over
a wide range of flow conditions, exhibiting decreasing carryover with decreasing
steam loading (flow rate).

An extensive literature search of technical papers, patents, and reports pertaining


to steam/water separator was conducted. The separator concepts and design features
were reviewed based on their performance. The available data on carryunder and
carryover performance of separators were defined in terms of system pressure,
steam loading, water level and water loading or quality. The separator design
features affecting the performance of primary separators are normally categorized
into one of four groups:

1. Vanes.

2. Pre-dryers.

3. Secondary separation devices.

4. Auxiliary parts to prevent entrainment.

Most separators reviewed were adequate for their particular application and a
few showed promise for further development for use in nuclear steam generators.
Three general types which warrant development to further optimize steam generator
design are:

1. The centrifugal vane, up-flow, down-flow separator because of high


steam loading capability.

2. The centrifugal vane, perforated jacket separator because of


excellent carryover performance over wide flow and exterior water
level ranges.

3. The axial vane separator because of high total loading capability.

The separators which are recommended for further development are compact in both
height and diameter so that they could be retrofitted to existing steam generators
as well as incorporated into future steam generators. A program outlining future
tests recommended for the development of these concepts is included in Appendix H
of this report.

Five milestones were set forth in this program:

1. Survey 1iterature.

2. Evaluate presently used steam separation equipment.

1-2
3. Evaluate proposed separator designs and separation systems.

4. Integrate presently used designs and/or proposed designs for


future development.

5. Recommend a test program for improved designs.

A brief description of each of these milestones is given in the Summary Section


of this report. The significant work performed in this program is summarized in
ten different areas. The highlights of these areas are contained in the Results
Section. The summary of each of these areas is contained in the Appendices
Section.

1. Literature survey

2. Presently used separator evaluation

3. Dryer design features and performance

4. Alternatives to two stage separation

5. Design features

6. Recommended separator concepts

7. Promising separator concepts

8. Recommended test program

9. Space/size 1 imitation evaluation

10. Hydraulic Evaluation

Possible increases in scope for a comprehensive separator development program are


briefly discussed in the Optional Studies and Test Section. These five studies/
tests are not mandatory for a separator development program, but they could be
desirable.

1. Screening test working fluid study

2. Sealing study and tests

3. Visualization study and test

4. Gravity separation tests, and

5. Dryer tests

1-3
Section 2

SUMMARY

The activities of this project consisted of surveying the open literature,


evaluating the performance and design features of separation equipment, and re
commending promising concepts and design features for further development and
testing. This project was conducted by Combustion Engineering, Inc. under a
jointly funded contract with Electric Power Research Institute. The work was
performed in accordance with Task Agreement $122-1 to meet the following four
objectives:

1. Develop a thorough understanding of the parameters which affect


separator and dryer performance and the related effects of separa­
tor performance on steam generator operation. This included an ex­
tensive design review of known separator size and configurations
including two stage separation, a study of separator hydraulics,
a review of steam generator/separator interactions, and a review of
dryer design and performance characteristics.

2. Determine the potential benefits to be derived from further separa­


tor development. This included an estimation of the benefits in
terms of improved steam generator operation, and the potential for
increased recirculation flow. The increased recirculation flow may
result directly from a reduced separator pressure loss or from
other steam generator design changes which would then require sepa­
rators with a higher flow loading capability.

3. Using the results of 1 and 2 above, identify a 1imited number of


promising improvements #n steam separator design which warrant
testing for further refinement and performance verification.

4. Recommend a test program, scope and schedule necessary for the


development and demonstration of an optimized separator design.

These objectives were accomplished in the following study topics or milestones

PROJECT MILESTONES

Five milestones were set forth in this program as a method of satisfying the
objectives. A discussion of these milestones follows:

2-1
Survey Literature

The 1iterature survey produced a large amount of information. The most pertinent
technical papers, patents, and reports are individually summarized in Appendix A.
The less pertinent literature, dealing with baffle plate separators, down-flow
gas/1iquid separators, dust separators, wire mesh demisters, and other surveys
of this nature, are not summarized, but merely 1isted in Appendix A for the readers
information.

Evaluate Presently Used Steam Separation Equipment

The separators and dryers used in Pressurized Water Reactor (PWR) steam generators
and Boiling Water Reactors (BWR) power systems were evaluated from the standpoint
of unique design features and operating performance. A hydraulic evaluation of
the separators pressure drop and carryunder on the recirculation flow in a steam
generator is presented in Appendix J. Separators designated as Design 1
(Curtis Wright), 2 (Westinghouse), 3 (General Electric), 4 (Kraftwerk Union), and
5 (Combustion Engineering) were evaluated and are included in Appendix B along
with sketches depicting the design features and plots of their performance.
Separators studied were of specific design and not meant to be generic examples
of a particular manufacturer. The specific carryover performance varies widely,
depending upon the type and design of each separator. All five of these separa­
tors, however, in combination with their respective dryer systems, show adequate
performance, though some have advantages over others for particular applications.
A general discussion of wire mesh, corrugated chevrons, and hooked-vane type
dryers is found in Appendix C along with sketches of some of the more commonly
used corrugated and hooked-vane chevrons.

Evaluate Proposed Separator Designs and Separation Systems

Most separators or dryers have design features which are similar and can be grouped
into four basic categories, as follows:

1. Vanes

2. Pre-dryers

3. Secondary separation devices

4. Auxiliary parts to prevent reentainment

2-2
In the development of a new separator concept, one or more of these designs fea­
tures could be incorporated into the separator concept to improve the performance.
A discussion of these features is given in Appendix E.

Three different types of extra stage separation devices were evaluated to deter­
mine if significant improvements in pressure drop, and/or carryunder and carryover
performance could be expected. These were:
1. Extra stage separators in series
2. Stacking of separators in parallel
3. Baffling in the riser for pre-separation

The second combination of separators showed the most promise for reducing the
pressure drop because the flow through each separator is reduced. With the proper
selection of separators stacked in parallel,it would also be possible to improve
carryover performance. The study of extra stage separation techniques is described
in detail in Appendix D.

Integrate Presently Used Designs and/or Proposed Designs for Future Development
Some of the best presently used and proposed separator design features were inte­
grated into three general concepts (centrifugal vane, up-flow, down-flow separa­
tor; centrifugal vane, perforated jacket separator; and axial flow separator)
which are recommended for further development. These three concepts are very
similar to three separators presently being used which are considered to be of the
best in the industry. Further development of these concepts is recommended because
of some of the 1 imitations of the presently used counterpart designs. These three
concepts are discussed in detail in Appendix F. In addition, ten other concepts
showing lesser promise (i.e., a lower probability of successful development and
therefore not recommended for further investigation) are described in Appendix G.

Recommended a Test Program for Improved Designs


A detailed discussion of a test program for the evaluation of the three recommended
concepts is included in Appendix H. Low pressure air/water screening tests are
recommended for initial development of the three concepts, followed by high pressure
steam/water proof tests to verify performance of the final three concepts with
optimized design features. The screening tests should be divided into two
phases.

2-3
1. Phase 1 should consist of comparative tests of similar design
features such as pre-dryers and secondary separation devices
followed by optimization of the design features of each con­
cept. For example, the ski rimer gaps, restrictor clearances, and
vane slopes will have to be determined for optimal performance as
a function of separator design.

2. Phase 2 should consist of comparative testing of the three final


concepts with the optimized design features.

After each component has been optimized with the screening tests, it should be
proof tested under high pressure steam and water conditions (800 and 1200 psi)
prototypical of an operating nuclear steam generator.

Freon 12, Freon 22, low pressure steam/water, and air/water were evaluated as the
working fluids to be used for the screening tests. The evaluation indicated
that Freon 22 modelled the fluid properties better than air/water, but is more
difficult to use. Air/water was chosen because it adequately models the water
droplets and steam bubbles in the separation process and much more experience
exists with air/water tests as a fluid for separator testing than with other
media evaluated.

2-4
Section 3

RESULTS

SEPARATORS

An extensive review of the separator 1iterature was conducted during this program
(Appendix A). The design features and performance of five separators presently
used were reviewed and presented (Appendix B). No conclusions are presented in
these two appendices. Several general results of separator operation and per­
formance are discussed.

Most nuclear steam generators have several separators supported by a separator


deck above the tube bundle. The region between the tube bundle and the separator
deck becomes the distribution plenum for the separators and the separator deck
collects the water separated from the steam by the separators. The water is routed
from the separator deck to the steam generator recirculation loop. Generally,
an increase in water level on the separator deck increases water carryover and
decreases steam carryunder. In many cases the flow loading (i.e., steam flow
rate and water flow rate) through the separators affects the performance of the
various separators differently. In Appendices B, F, and G the performance
characteristics of various separators are presented. In selecting the proper
separation equipment for a given application, consideration must be given to the
steam and water flow rate distribution from the plenum to the entrance of the
separators and to the water level profile which exists on the separator deck to
ensure that the equipment selected will operate properly under the predicted
conditions.

DRYERS

Several types of dryers were reviewed because of the interrelationship with overall
steam separation system design in a steam generator application. The baffle dryers
in which the flow traverses in the horizontal direction had significantly better
performance than those baffle dryers in which the flow passed through the vertical
direction. By inclining baffles up to about 30° the performance could be improved
since there is an increase in the effective flow area. The two types of dryers

3-1
commonly used in nuclear steam generators today are those with corrugated chevrons
and those with hooked-vane chevrons. The corrugated chevrons are compact in size
and extract a smal1 percentage of moisture (less than 10%). In comparison, the
hooked-vane dryer occupies considerable volume, but extracts up to 30% of the
moisture. Dryers are able to extract more moisture at lower inlet velocities.
When the velocity exceeds a critical threshold velocity, carryover from dryers in­
creases substantially. In selecting a dryer design then, consideration must be
given to the inlet conditions (percentage of moisture and inlet velocity).

The flow distribution to each dryer and across the flow area of each dryer should
be nearly uniform to avoid local peaks which may exceed the critical velocity.
There are several designs that have been used to ensure more uniform flow distribu­
tion to the dryers.

1. Orfficing of Selected Separators - In U-tube steam generators,


there is more steam generation on the hot leg side as opposed to
the cold leg side. The steam separators, in some cases, do not
act as an adequate flow distributor. In these cases, added
resistance to selected separators on the hot leg side has been
used to ensure more uniform flow through the separators and dryers.

2. Two tiers of dryers - If the dryer height is too large (greater


than 20"), then it is possible that the moisture as it drains
down the chevrons wi11 fill the lower portion of the dryer,
resulting in massive carryover. One design to prevent this is
to use two tiers of dryers instead of one tall dryer.

3. Flow distribution baffling - When two tiers of dryers are used


with parallel flow paths there is a tendency for the flow to go
through the lower tier of dryers. Flow distribution baffling
(perforated plates) have been used to obtain more uniform flow
distribution to the lower and upper tiers of dryers.

HYDRAULIC EVALUATION

In general, higher circulating flow rates are considered advantageous in in­


creasing reliability of PWR steam generators. A Combustion Engineering, Inc.,
hydraulic computer program "CRIBE" was used to determine the effect separator
flow resistance and carryunder had on the circulating flow rate of a steam
generator. The separator shock loss factor was varied for a Combustion
Engineering non-economizer steam generator, a Westinghouse non-economizer steam
generator, and a Combustion Engineering economizer steam generator. The flow
resistance of the separators in the recirculation loop was found to be about 20%
of the total. The results showed that substantial reduction in separator pressure
drop does not significantly affect the overal1 circulation flow rate. For example,

3-2
a Combustion Engineering non-economizer type steam generator with a separator
shock loss factor of 3.0 at full power has a circulation ratio (CR) of 3.9 (cir­
culation ratio is the ratio of total flow to steam flow that enters the separators).
Reducing the separator shock loss factor to 1.5 only increases circulation ratio
to 4.2. That is cutting the pressure drop through the separator in half increases
the CR only approximately 10%.

Steam carryunder decreases the density in the downcomer and hence decreases the
driving head to cause recirculation. The "CRIBE" results for a Combustion
Engineering non-economizer steam generator indicated that for ful1 power, there
is negligible change in CR with up to 6% carryunder. At 50% power, the decrease
in CR becomes more pronounced, decreasing the CR from 8.45 at no carryunder to
8.0 at 6% carryunder. For an economizer steam generator, the full load CR de­
creased from 3.2 at no carryunder to 2.5 at 6% carryunder. At 50% load, CR de­
creased from 9.3 at no carryunder to 7.9 at 6% carryunder.

The conditions evaluated during this study were typica1 of steam generator
operation but may not be the same as those used in the final evaluation bythe
steam generator supplier. As a consequence the circulation ratio quoted herein
should be considered appropriate for the comparison of effects intended in this
study.

PRIMARY SEPARATOR SPACE/SIZE AND PERFORMANCE EVALUATION

A space/size and performance evaluation was conducted using the steam separators
designated as 1 (Curtis Wright), 2 (Westinghouse), 3 (General Electric), 4 (Kraft­
werk Union), and 5 (Combustion Engineering), (Appendix B). The steam generators
used in this evaluation were designated as 1 (Combustion Engineering), 2 (Westing­
house) , and 3 (Kraftwerk Union), (Appendix I). The evaluation conducted was to
determine if each of the five separators could be installed into the three steam
generators and how the performance would be affected with the various combinations
of separators and dryers. This evaluation is included in Appendix I. Two separa­
tor combinations showed promise. If the support deck of steam generator 2 could
be modified to accept separator 5, then steam quality entering the dryers should
be less than 0.25%. Steam separator 3 fits through the manways of al 1 three
steam generators. This separator, with further development, has the potential
of allowing the steam generator to be fabricated as a straight shell.

3-3
ALTERNATIVES TO TWO STAGE SEPARATION

Alternatives to the normal practice of two stage separation (separators followed


by dryers) were investigated, and the results are presented in Appendix D. Five
schemes were evaluated to determine if pressure drop in the circulating loop of
the steam generator could be decreased or carryover could be decreased without
increasing carryunder.

Extra Stage Separators in Series

A second stage of separators between the first stage of separators and the dryers
is not recommended. For most separators used in a nuclear application, the exit
moisture is less than 10%. The second stage of separators evaluated did not
operate effectively for inlet quality greater than 90%.

Additional Stage of Dryers in Parallel

An extra stage of dryers in parallel has merit. The flow area would be increased,
thus decreasing the through velocity and reducing carryover. Because the emphasis
of this program was on separators, not dryers, this type design improvement was
not pursued. In designing a two stage dryer systern in pa rail el, one should pro­
vide for near uniform distribution to the dryers and drainage from the dryers.

Extra Stage of Dryers in Series

An extra stage of dryers in series is not recommended. If excess carryover from


the first stage of dryers were occurring, the through velocity would be greater
than the critical velocity for that dryer design. Assuming equal surface in the
second stage of dryers, this condition then would also be present in the second
stage of dryers, resulting in no significant decreases in carryover. (The criti­
cal velocity is that velocity above which carryover begins increasing drastically.)

Stacked Separators in Parallel

Stacking of separators in parallel has potential for lower carryover for several
separator designs such as the Curtis Wright Design 1 separator and the Combustion
Engineering Design 5 separator. The Design 1 and Design 5 separators have the
characteristic that for decreasing steam flow rate, carryover decreases. Stacking
of the Combustion Engineering separator in parallel is recommended for development
and described in Appendix F.

3-4
Baffling above Tube Bundle and Below Separators

Baffling above the tube bundle and below the separators was considered because it
had the potential of increasing the circulation ratio. Baffling would perform
gross separation at a low pressure drop, and the troughs of the baffle would
channel the water to the downcomer. This scheme is not recommended because the
evaluation previously described shows that a significant improvement in CR can
not be obtained since the separators only contributed to about 20$ of the pressure
loss in the circulating loop of the steam generator.

DESIGN FEATURES

In the design of a separator, one or more of the following design features are
usually incorporated.

1. Vanes - Vanes are probably the most important feature since they
induce the centrifugal motion to the mixture which causes separa­
tion of the steam and water and create the majority of the pres­
sure drop.

2. Pre-dryers(Secondary Separators) - Pre-dryers are an integral part


of the separator located at the separator steam exit. A pre-dryer
can be wire mesh contained in a perforated cylinder, deflector
plate, vane arrangement, or others. Pre-dryers are normally used
to extract moisture and/or distribute the steam flow.

3. Secondary Separation Devices - Secondary separation devices are


usually baffling or a vane arrangement that is 1ocated in the
downcomer of the separator. The steam carried under in the
exiting water is separated by these devices and allowed to flow to
the dryers.

4. Auxiliary Parts to Prevent Reentrainment- Devices such as skim­


mers, restrictors and perforations are sometimes used to prevent
the reentrainment of water droplets in the exiting steam flow.

In many cases the contribution of an individual design feature to the separator's


operation or performance is not known since testing costs preclude their being
studied separately. This is one of the reasons that screening tests are re­
commended for any follow-on test program. A detailed discussion of design
features is presented in Appendix E. No conclusions can be made with respect
to the desirability of each of the above design features as they are only
individual contributors to the overall performance of a given separator.

RECOMMENDED SEPARATORS

The primary objective of an improved separator design is to obtain a separator


which will accept a wide range of flow loadings while exhibiting a very low

3-5
carryover combined with acceptable carryunder performance. A second objective
is to have the design compact enough so that it can be easily retrofitted to a
typical steam generator with a 16" or larger manway.

Carryover warranty is generally 0.25% exit moisture for nuclear steam generators.
The conventional separation system for most nuclear steam generators is a first
stage of separators which does the bulk of the separation and a second stage of
dryers which performs the fine separation. Usually both stages of separation
equipment are required in series to remove the moisture so that outlet steam
quality is greater than 99.75%.

The separators evaluated in this program were grouped into three categories based
upon their degree of promise for further development. The first category of sepa­
rators showed little promise based on available performance information and were
eliminated from consideration for further development based upon engineering
judgment. These were mostly used in applications (fossil, petroleum industry,
dust separators, etc.) where the environment was quite different from that of a
nuclear steam generator.

The second category of separators are those which have some merit with respect
to one or more operational or performance parameters, but do not show enough
overall promise to warrant further development efforts. Sketches of these
separators, along with performance curves and specific reasons for not choosing
them for further development, are presented in Appendix G. This category is com­
prised of the following separator designs.

1. Separator with Hollow Vanes in Annulus (Russian Separator)

* 2. Curved Arm Separator (Curtis Wright Separator)

* 3. Turbo Separator with Tangential Ports (Westinghouse Separator)

4. Separator with Converging Shell (Foster Wheeler Separator)

5. Early Axial Flow Separators (General Electric Separators)

6. Radial Vane Separator (Separator Developed for Atomic Energy


Commission)

7. Turbo Separator (Combustion Engineering Separator)

8. Downflow Separator (Babcock and Wilcox Separator)

9. BWR Downflow Separator (AEG Separator and a Allis Chalmers Separa­


tor)
10. Axial Flow Venturi Separator (Separator Proposed by Prins in
Thesis)

* Presently used separators described in Appendix B.

The third category of separators are three which are recommended for further deve­
lopment. Each of these three concepts, as they are presently used in industry
application, is among those discussed in Appendix B. Sketches of these type
separator designs, along with performance curves and design modifications, are
presented in Appendix F. Again these are specific designs and not a generic
manufacturer separator. A synopsis of the three separators recommended for further
development follows:

Centrifugal Vane, Up-flow, Down-flow Separator (Figure 1A)

Kraftwerk Union developed a modification of their separator (Figure B-21) which


has the highest steam flow rate loading per unit area of the separators evaluated.
For loadings of steam flow rate of 84000 Lbm/hr and water flow rate of 220000 Lbm/hr
and water level of 40 inches, the carryover and carryunder performance of this
compact separator (Figure F-l) is about 2% and 0.1%, respectively. The major draw­
back to the compact separator is that a decrease in water level results in a drastic
increase in carryunder and an increase in water flow rate from the design conditions
results in a sharp increase in carryover.

The goal of developing a centrifugal vane, up-flow, down-flow separator is two­


fold. First, the separator should be compact enough to be retrofitted into
existing steam generators. Second, it should have a high allowable steam flow
rate loading per area and operate over a wide range of water levels and water
flow rates with carryover of about 2% and carryunder of about 0.5%. Brandes
(Reference 5 of Appendix B) describes a similitude analysis for the centrifugal
vane, up-flow, down-flow separator (Figure 1 A ). A less sensitive design may be
perfected through the use of this similitude analysis in conjunction with verifi­
cation testing.

Centrifugal Vane, Perforated Jacket Separator (Figure 1B )

This separator (Combustion Engineering) is recommended for further development


because its carryover is in the 0.25% range at low steam loading. Design and
arrangement modifications recommended for development testing with this separator
are alternative perforation geometries and stacked separators in parallel to decrease
the steam loading per separator. Based on an evaluation of the performance data

3-7
available, these modifications should produce even better carryover performance.
The centrifugal vane, perforated jacket separator recommended for testing in the
stacked arrangement is shown in Figure IB.

Axial Flow Separator (Figure 1C)

This separator (Figure B-10, General Electric) has the highest total (steam and
water) flow loading per unit area of those evaluated. The purpose of further
development of this separator is to reduce the height for retrofitting and to
modify the restrictor clearance to accept an inlet quality of 25%. Most of the
nuclear steam generator separators operate with an inlet quality of approximately
25%, but this separator is currently designed for and used with an inlet quality
of 10 to 15%. A modification to this separator to be tested is shown in Figure
1C.

i Screen
Assembly

Skimmer
Upper Exit Nozzle
Riser Straightening
Vanes

Downcomer

Downcomer Skimmer

Straightening
Vanes
GJ
Vanes
ro_______
<£>
Perforations

Downcomer

Restrictor
Vanes

Lower
Riser
Vanes

A. Centrifugal Vane, Up-flow, Down-flow B. Centrifugal Vane, Perforated C. Axial Flow Separator
Separator. Jacket Separator

Figure 1. Recommended Separators.


Recommended Test Program

Appendix H is a detailed discussion of the recommended test program for the


evaluation and development of the three separators discussed above. Basically
there are two recommended series of tests: low pressure air/water screening
tests followed by high pressure steam/water proof tests. This sequence of deve­
lopment was envisioned as a method of economy. The screening tests are relatively
inexpensive and can be used to provide comparison and optimization of the variation
in design features associated with each separator design. This will provide an
optimized design before the relatively expensive proof tests are performed to
verify performance under prototypical steam generator operating conditions.

There should be several benefits from these tests besides developing a improved
separator.

1. One benefit should be in developing a better correlation between


low pressure air/water tests and high pressure steam/water tests.
Because of the difference in properties of the fluids and thus a
variation in flow dynamics, a correlation between air/water and
high pressure steam/water for separator operation is needed. Con­
ducting a large number of tests using high pressure steam/water may
be prohibitive because of the high fuel costs. In this proposed
separator developmental program, air/water is recommended as the
working fluid for the bulk of the tests. High pressure steam/water
would only be used for the proof tests of the three developed sepa­
rators . With air/water test data and high pressure steam/water
test data available for the same separator, it may be possible to
determine a correlation between the low pressure air/water data and
high pressure steam/water data.

2. A second benefit should be derived from using the data to check the
similitude analysis, which would be necessary for any testing of a
scaled-down separator. Two sizes of the centrifugal vane, up-flow,
down-flow separator (Figure 1A) should be tested in addition to
variations in the design dimensions. Brandes in Reference 5 of
Appendix B describes a similitude analysis for this type separator.
If Brandes similitude analysis or a variation of his similitude
analysis proved valid, then model-prototype separator development
could be performed with more confidence.

3. A third benefit should be the screening test results on various


design features. Various pre-dryer and secondary separation de­
vices would be tested on a comparison basis. Their operation would
be observed in air/water screening tests. Variations of the vane
dimensions, restrictor clearances and skimmer clearances of the
centrifugal vane, up-flow, down-flow separator (Figure 1A) and the
axial flow separator (Figure 1C) would be tested. These tests
should allow one to make better judgments on the benefits of in­
dividual design features to be incorporated into new or modified
separators.

3-10
Section 4

OPTIONAL STUDIES AND TESTS

Appendix H presents the minimum recommended test program which is required for
further development of the three selected separator designs. This section of the
report presents optional studies and tests which would be highly desirable to
enhance the general knowledge for separator development, but are not considered
mandatory to the further development of a specific separator concept.

The optional studies and tests presented in this section were given consideration
as being part of the recommended test program, but were eliminated to produce an
economical minimum development program for improving primary separator performance.
These options are summarized as follows:

SCREENING TEST WORKING FLUID STUDY

A preliminary evaluation was performed on separator testing fluids and it is re­


commended that low pressure air/water screening tests for the comparison and
optimization of discrete separator design features be used in the test program.
A more thorough study which includes a wider variety of fluids might yield a
working fluid which would provide a better correlation between a screening test
and a high pressure steam/water test than does air/water testing. The main
purpose of this type study is to supply characteristic parameters that could be
used in mathematical model 1ing of separators.

SCALING STUDY AND TESTS

Scaling of a separator size, without changing the performance characteristics,


would have two advantages; it would permit increasing or decreasing the size of
a separator for a specific steam generator application and it would permit more
economical testing of a multiple cluster of separators to determine the effects
of interactions among separators under test conditions prototypical of an
operating steam generator. Sealing down the separator would be required to test
interaction effects because most existing test drums are only large enough to
accomodate one ful1 size separator. A sealing study and test is required to prove
that seal ing can be accomplished without affecting separator performance charac­
ter i sties.

4-1
VISUALIZATION STUDY AND TEST

Presently the industry is at the point in separator development where an approxi­


mation of the hydraulics (bubble and droplet size) within a separator under proto­
typical steam/water conditions must be determined in order to gain significantly
in analytical design capability. A visualization study and test program would
be necessary to develop the techniques required for observation of the flow
phenomena occurring inside a separator. Fiber optics and/or view ports could
be used.

GRAVITY SEPARATION TESTS

Performing gravity separation tests, which model actual operation in a PWR steam
generator riser (the region above the tube bundle and below the separator support
deck), should be considered to better characterize separation in this region of a
steam generator. The actual flow conditions at the entrance to separators could
be more accurately predicted and modelled in test programs if the gravity separa­
tion in the riser region was more fully understood.

DRYER TESTS

Since dryers pi ay such an important role for steam/water separation in a steam


generator, it would be useful to test the various dryer designs in the same test
loop under the same conditions. Past dryer tests have been so widely diversified
that it is difficult to produce a comparactive performance evaluation. Utilizing
the data developed, a designer would then have the flexibility of choosing a dryer
based on the needs of his specific application whether it be space requirements
or meeting rigid dryer performance specifications.

4-2
APPENDIX A

LITERATURE SURVEY
ILLUSTRATIONS

Figure Page
A-l Moisture Entrained 15 Inches Above Interface in 4-Inch
Diameter Test Section A-2
A-2 Nested Vortex Separator A-3
A-3 Candidate Separators A-10
A-4 Developed Separator A-10
A-5 Diffuser A-12
A-6 Radial Arm Separator A-13
A-7 Axial Flow Horizontal Separator A-15
A-8 Critical Liquid Downflow Velocity for 5 Volume
Percent Carryunder A-18
A-9 Separator for Nuclear Power PI ant A-20
A-10 Separation Apparatus A-22
A-11 Horizontal Separator A-23
A-12 Axial Flow Separator with Castellations A-25
A-13 Flow Distribution Apparatus A-27
A-14 Gravity Separation Test Section A-31
A-15 Curved Arm Separator A-33
A-16 Test Drum Internals-High Pressure Tests A-34
A-17 Effect of Submergence on Separation Height A-35
A-18 Effect of Steam Flow on Separation Height A-36
A-19 Effect of Water Flow on Separation Height A-37
A-20 Effect of Pressure on Separation A-38
A-21 Water Steam Separator A-39
A-22 Tangential and Archimedes Spiral Separators A-41
A-23 Circulation Loop A-44
A-24 Wetness of 2 Euroform Water Separators A-52
A-25 Vane Elements A-52
A-26 Separation Devices A-57

A-i
Figure Page
A-27 Arc Separator A-58
A-28 Axial Flow Separator with Secondary Vanes A-60
A-29 Test Separator A-63
A-30 Downflow Separator A-66
A-31 Upcomer Separator A-67
A-32 Hooked Crossover A-68
A-33 Horizontal Separator A-69
A-34 Centrifugal Dryer A-72
A-35 Centrifugal Dryer Test Results A-73
A-36 Chevron Type Dryer A-74
A-37 Operating Characteristics of Chevron Type Dryer A-75
A-38 Methods of Fabricating Mesh Type Dryer A-76
A-39 Operating Characteristics of a 5-LB-PER-CU-FT,
6-Inch-Thick Mesh Type Dryer A-77
A-40 Moisture Carryover Tests A-78
A-41 Mechanisms Governing Liquid Entrainment A-80
A-42 Entrainment as a Function of the Reduced (Superficial)
Vapor Velocity for Water at 185 ATM and for the Height
of the Vapor Dome H=0.465 m. A-81

A-ii
Section 1

SUMMARIES

1. Moisture De-Entrainment Tests in Two and Four-Inch Diameter Test Sections


ACNP 5921, November 15, 1959.

Carryover was measured in a four inch diameter test section. The steam f1 owed
through a perforated plate and a fine-mesh screen at the entrance of the test
section. Figure A-l shows carryover as a function of the steam release rate
for pressures of 300, 400, 500 and 600 psig at 15" above the interface.

2. Babcock and Wilcox Company


Vapor-Liquid Separating Apparatus
British Patent 817,650

The apparatus described in this patent is a downcomer tangential flow separator.


See Appendix G, Number 8. The vapor-1iquid enters through a tangential nozzle
in the upper section of the separator. The curvature of the barrel creates
the centrifugal force for separation. The separating effectiveness is enhanced
by a skimmer baffle which divides the 1iquid from the steam. The steam exits
upward through a discharge nozzle and pressure distributor. An end baffle below
the tangential nozzle prevents the steam from flowing downward. An annular
passageway is provided at the end of this baffle for the water to flow. Im­
peller vanes in this passageway give the water more of a downward velocity.
The water exits the bottom of the separator through spouts.

3. Bailey
Nested Vortex Separator
U.S. Patent 3360908

The nested vortex separator (Figure A-2) is a plurality of vortex elements


nested in a common casing. This separator has application in large pipelines.
The length-diameter ratio of the separator elements allows for a higher capa­
city.

A-l
600 PSIG

500 PSIG

400 PSIG

300 PSIG

0 0.5 1.0 1.5 2.0


R/P 9
Note: R=Steam Release Rate in Test Section (LB/HR-FT^)
2
P=Pathfinder Design Steam Release Rate (LB/HR-FT 6650 LB/HR-FT

0 3325 .......6650.... 997 5 “'13300...........1662 5


STEAM RELEASE RATE (LBM/HR-FT2)

Figure A-l. Moisture Entrained 15 Inches Above Interface in 4- Inch


Diameter Test Section.

A-2
Figure A-2. Nested Vortex Separator.

A-3
4. Bartolomei and Alkhutor
Determination of the True Vapor Content When There is Bubbling in the
Stab!ization Section
Teploenergetika, Vol. 14, No. 12, 1967

For bubbling steam through a water pool, Bartolomei in this paper correlated
several groups of experimental data to obtain a relation for volumetric vapor
content.

a = 1 - exp (-ca^)

Where

c = 2.90

and

0.15 .

d =i^7T

a = Volumetric Vapor Content


c = Coeefficient
y^= Liquid Viscosity
vig = Gas Viscosity
p = Gas Density
a = Density Difference Between Phases
p
a= Surface Tension
g = Acceleration Due to Gravity
D = Diameter of Apparatus in which Bubbles are Formed

A-4
5. Batishchev
The Operation of Horizontal Chevron-Type Separators
Teploenergetika, Vol, 15, No. 8, 1968

The author investigates and reports on the operation and effectiveness of


horizontal baffle separators in this paper. He contends that there is a mis­
conception in that the deterioration of carryover performance is associated
with choking or flooding. The deterioration of carryover performance occurs
when the steam velocity is high enough to prevent droplets from adhering to
the plate surface, thus carryiny the moisture through the baffle separator.
The author reviews the equations and plots from other investigators on the
performance of horizontal baffles. He also investigates the conditions for
the formation of the water droplets. The conclusion of his study is that
vertical baffle separators are preferred over horizontal baffle separators.

6. Bennett and Kondic


Momentum Flux Model for Liquid-Vapor Separation
1977 National Heat Transfer Conference

A momentum flux model is presented to calculate the efficiency due to centri­


fugal and gravity separation. For centrifugal separation, two parameters need
to be determined before the efficiency can be completed:

1. Momentum flux difference (calculation)

2. Diameter - Viscosity characteristic group (experimental data)

The efficiency is the ratio of a difference term (inlet water flow rate-water
carryover) and inlet water flow rate. The equation for the water carryover is
equal to 1/2 the inlet water flow rate times the ratio of the exit orifice
area and the area within the center of mass of the liquid.

7. Blum, Sievers and Wagner


Steam Generator Design and Development Capability
Curtis Wright Report, June 1977

Design information and test data are given in this report for a model curved
arm separator operating at about 350 psi. Below a steam flow of 20000 Lb/hr
and a circulation ratio of 6, this separator has excellent performance. See
Appendix B, Number 1.

A-5
8. Boiling Reactors: Steam-Water Separation
Section VIII Power Reactor Technology
Vol. 6, No. 2, 1963

This paper is a summary of separator technology. The author(s) give the


following relation for downcomer slip ratio. High pressure steam/water data
as well as air/water data was used in determining this relation.

Where

^g^L = Ratio of gas and liquid velocities


Fr = Froude umber
x = Weight fraction of gas
= Density of liquid
Pg = Density of gas

9. Bouecke and Schucktanz


Two Phase Flow Separation in U-Tube Steam Generators
1977 National Heat Transfer Conference

In this paper, the authors give high pressure test results for the upflow-
downflow separator and compact separator. See Appendix B, Design 4 and
Appendix F, Number 1.

10. Brandes
Determination of the Performance of Steam Separator with the Aid of the
Laws of Similitude
Kerntechnik, Vol. 14, No. 5, 1972

Similitude analysis was used by the author to model the upflow-downflow


separator (See Figure F-4 and F-5) based on pressure loss and separation
factor. For pressure loss testing, seven characteristic groups were defined.
Four of these were geometric ratios, the fifth is the ratio of vane diameter
to water level. The sixth is the pressure loss coefficient, and the seventh
is a "Froude" number. For different separator geometries, the author found
a correlation existed for pressure 1oss coefficient versus Froude number.
The Froude number was a function of velocity through the vanes, curvature
of the vanes, vane exit angle, and mean length between two bubbles.

A-6
Similar characteristic groups were used for the similitude analysis based on
separation. Of the nine groups used, four were based on geometry, the fifth
was water level divided by vane diameter, the sixth was mean distance between
bubbles per vane diameter, the seventh, eighth, and ninth were a Weber number,
Froude number, and separation factor, respectively. For various geometries
of the separator, the author found a correlation of separation factor versus
a characteristic parameter. This parameter is a function of Froude number,
Weber number, bubble diameter, vane diameter, vane curvature, and exit vane
angle.

The author indicates that by using either of the similitude analyes, it was
possible to scale this separator.

11. Brandes
Experimental Investigation of Centrifugal Separators - Part 1
Energie and Technik, Vol. 23, 1971

By theoretical considerations, the pressure drop across a upflow-downf1ow


separator (See Figures F-4 and F-5) was analyzed. By mechanics of similitude,
the characteristic factors for pressure loss were determined. For the separa­
tion process, these conditions were evaluated: centrifugal force, gravitational
force, and capillary force. Likewise, the characteristic parameters for the
separation process were determined.

The test setup and instrumentation were also described. They 1isted the varia­
tions of the upflow-downflow separator tested.

12. Brandes
Experimental Investigators of Centrifugal Separators - Part 2
Energie and Technik, Vol. 23, 1970

The experimental air/water test results for the upflow-downflow separator


(See Figures F-4 and F-5) are delineated in this paper. The results indicate
that the separator pressure drop increases with the following:

1. Decreasing overal1 diameter

2. Increasing curvature of vanes

3. Increasing discharge angle

4. Increasing height

5. Increasing hub diameter

A-7
The author found that the ratio of diameter of center steam exit tube to
separator diameter significantly affected the performance. Too-large a ratio
would result in incomplete separation in the downcomer. The author's tests
indicated that a water column was needed above the blade channel to stablize
the flow.

For the centrifugal vanes, the author found that the optimal exit angle was
about 42° for a minimum pressure drop. Too small a deflection results in broad
boundary 1ayers and too large a deflector leads to a larger shock loss. The-
author also found that the flow behaved differently for air/water mixtures
through the lattice (vanes) than single phase air.

A presentation of the test results by the theoretically determined characteris­


tics is also made.

13. Burley
Performance of Internal Steam Separati on System in Boiling Water Reactor
ASME Paper 69-WA/NE-24

Burley discussed the tests made on the separator/dryer system during the KRB
startup. At rated conditions, carryover from the dryers was 1% and carryover
from the separators was 10%. Carryunder was 0.6% by weight at ful1 power
conditions. At a very low water level, the profile varied 6".

14. Burley, Kudirka, and Moen


Performance of Internal Steam Separation System in Large Boiling Water
Reactor Plants
ASME Paper 72-PWR-6

Carryover and carryunder measurements were taken to 10 BWR's during operation.


Sodium-24 trace techniques were used to measure carryover, and a heat balance
method was used to measure carryunder. Separators similar to the ones shown
in Appendix G, Number 5, were used.

The water level tests carried out indicated 1ittle variation in carryover-
carryunder from center to periphery. The carryover results indicated almost
uniform distribution.

15. Cagagrandi, Bertoletti, and Perana


Separation at High Quality
CISE, January 1963

Five separators were screen tested using air and water at the working fluids.
Four of these separators (Figure A-3) had porous walls with different vanes

A-8
and steam outlets. The authors chose to develop the separator shown in
Figure A-4. They found that whether the width of the vanes covered 2/3 or
the whole cross section made of little difference in performance. The vane
pitch was optimized at 45°. The poriosity of the wall had a significant
influence on this separator's performance. Successive layers of punched
steel sheet (up to a total of three) improved the performance. The shroud
for these tests was 3.9" and the diameters of the perforated sheets were 2.1",
2.2" and 2.4".

The developed candidate (Figure A-4) was tested using high pressure steam water.
These tests confirmed the screen tests. At a pressure of 700 psia, the
efficiency was 100% for flow rate per area of 615, 513, 410 and 308 Lbm/ft^S
up to inlet qualities of 40%. Above this quality, the separator efficiency
began to decrease.

16. Carpenter and Othmer


Entrainment Removal by a Wire-Mesh Separator
A.I.Ch.E. Journal, Vol. 1, No. 4, December 1955

Four inches thick wire mesh was tested for separator's properties. The carry­
over efficiency increased from 80% at a flow velocity of 2 ft/sec to near 100%
at 9 ft/sec to 23 ft/sec but decreased for higher velocities. Carpenter re­
commended using the followig equation to calculate the separation efficiency.

E = 1 - (1 - Em/c)N

N = number of 1ayers
C = parameters that is characteristic of separator
Em = efficiency of single fiber

17. Cochran
Axial Flow Vapor-Liquid Separator
U.S. Patent 3216182

This patent is for an axial flow vapor/1iquid separator. See Concept 5 of


Appendix G. The vanes are axial with a conical central hub which accelerates
the flow. There is a wire mesh pre-dryer to prevent excess moisture from
entering the dryers. A secondary separator device is installed at the top of
the secondary discharge channel to remove carryunder.

A-9
Figure A-3. Candidate Separators.
3

Figure A-4. Developed Separator.

A-10
Several of the design features were optimized. A vortex tube end gap of 1"
and inner tube overlap length of 0.5" minimized carryover. The author gave
empirical equations for establishing the separator diameter and vortex tube
length. He indicated that separators of this type could be scaled in size
by using a modified Froude number and volumetric flow rates of the model and
prototype.

18. Cochran
Pressure Recovery Axial Flow Vapor-Liquid Separator
U.S. Patent 3329130 and British Patent No. 1157877

This patent describes an axial flow separator with a diffuser (Figure G-ll)
in the inner downcomer passage. The diffuser blades (Figure A-5) reduce the
velocity of the 1iquid and convert some of the kinetic energy to pressure head.
A restriction is placed in the annul us below the diffuser blades to create a
back pressure to prevent flow separation or flashing.

19. Daman, Vail, B1izard and Phillips


Vapor-Liquid Separator
U.S. Patent 3296779

This patent describes a separator (Figure A-6) which operates as follows. The
steam/water mixture enters through the riser. The top piates turns the flow
through the slots into the curve arm channels where the separation occurs. The
shape of the vortex is controlled by the flow directing vanes in the curved
arms. The separated steam flows upward and the water downward. The swirl is
eliminated from the flow as it exits the downcomer by vortex eliminator vanes.

20. Dement1ev and Malinin


Experimental Investigation of Centrifugal Steam/Water Separator
Teploenergetika, Vol. 21, No. 9, 1974

Three model separators were tested at 290 psi in a steam/water test facility.
The designs had one, two and three skimmers respectively. Each of the separa­
tors had vanes with an exit angle of 25° at the boss and 55° at the wall.

The performance of the three separators were similar in several respects.


Carryunder, carryover and pressure drop increased with load. The resistance
coefficient for total separator pressure drop was fairly constant, independent
of load. The experimenters noticed that there was some fluctuation of pressure
drop under steady state conditions. This was attributed to a siug flow
regime of the mixture at the inlet.

A-ll
A-12
TOP PLATE

CURVED ARM

RISER

Figure A-6. Radial arm separator.

A-13
21. Dement1ev, Moirer, Skachek, and Krivov
Experimental Investigation of Centrifugal Steam Water Separator
Teploenergetica, Vol. 19, No. 3, 1972

Air/water tests and low pressure steam/water tests were conducted in this
separator developmental program. The results were extrapolated to 75 psia.
The initial separator tested was a 2.8" diameter pipe with round steam exit
holes, and the entrance contained four spiral vanes. Using air/water, the
vane design was optimized. The number of vanes was increased to six and the
outlet angle set at 55° with 15° overlap. In the next series of tests, the
developers investigated optimizing the shape and size of the holes. They
found the following:

1. The holes should begin in a section where separation of phases has


been completed.

2. The flow area of holes should be 2.5 to 3 times larger than the
diameter of the pipe.

3. Overflow ports are superior to round holes.

To prevent interaction between groups of separators and possibly prevent


deterioration in performance, each of these separators was placed in a cy­
lindrical casing. Further tests indicated that 7.1" was an adequate choice
for the casing diameter. Re-entrainment occurred in the form of an air/water
emulsion in the downcomer. Vertical guide vanes in the annul us (downcomer)
along the entire length of the overflow ports were used to remedy this pro­
blem. This led toward secondary separation; the water was pressed to the
casing wall and flowed downward. The air concentrated around the pipe and
escaped through the vane channels.

For steam/water tests the pipe diameter was increased to 3.9" and the casing
diameter to 7.5". The low pressure tests were at 160, 305 and 520 psi with
a steam flow rate of 660 to 18740 Lbm/hr with a water flow rate of 44100 to
154350 Lbm/hr. Interestingly, the carryover test data for three pressures
correlated as a function of the kinetic energy of the two phase flow. The
water level was probably at some constant height below the overflow ports.
See Appendix G, Number 1.

22. Device for Water Separation


British Patent 1259901

The axial flow horizontal separator (Figure A-7) described in this patent has
several unique features. The mixture flows through an agglomerator positioned
at a 45° angle to maximize the flow area. The flow enters five tubes with

A-14
SI i ts

Figure A-7. Axial Flow Horizontal Separator.

A-15
inclined vanes. The seven separate tubes allow the overall length to be
minimized. Strands of water collect in the guide ribs and are conveyed to
the drain holes. The film of water that collects between the guide vanes is
eliminated by narrow si its.

23. Fitzsimmons, Hughes and Moen


Gas-Liquid Separator
U.S. Patent No. 3603062 and British Patent No. 1259711

Two unique features of this axial flow separator are the pre-dryer and the
secondary separation device. See Concept 5 of Appendix G. The pre-dryer
is a vortex tube with a skimmer near the exit. It has a small diameter
about the same as that of the separator exit nozzle. The reduced diameter
increases the velocity of the water vortex and increases the film thickness
so that it can be skimmed off. The secondary separation device is a series
of alternate perforated and non-perforated baffles. A vortex is created
allowing the steam to move to the inside through the perforated baffle out
an exit port.

24. Gammon
History and Development of Water Separators
Filtration and Separation July/August 1973

The equipment described in this paper (filter separators) has the primary
objective of using Stokes Law to do most of the separating work. These
devices make large drops out of small drops. The settling velocity
increases as the square of the drop diameter per Stokes Law.

These filter separators remove water and dirt from fuel. The first stage
is a coalescing element which stops dirt at the inlet, and water is agglom­
erated as it passes out the exit face. The second stage filter is impervious
to water so only the hydrocarbon phase flows through the element.

25. Gardner, Crow, and Neller


Carryunder Performance of Drums in High Pressure Circulation Boilers
Proc. Instr. Mech. Engr., Vol. 187, 1973

The authors investigated carryunder in a high pressure test drum. Freon 12


was employed to test a steam separator and conduct natural separation tests
(gas bubbles rising in 1 iquid pool). The following modelling rules were
used to relate Freon 12 with high pressure steam/water:

A-16
1. Vapor to liquid inflow the same.
2. Identical fl ow patterns (equality of Froude number achieved this).

2
3. ,..P....P equated for model and prototype.
ApC)

p = Liquid density
y = Liquid viscosity
a = Surface tension
Ap = Density difference between phases

For 5 percent carryunder (by volume), Figure A-8 shows the critical down­
flow velocity with respect to water level. The authors indicated that if
the 1iquid velocity exceeds 0.3 ft/s to 1.0 ft/s vapor will be carried
under.

26. Garner, Ellis, Lacey


The Size Distribution and Entrainment of Droplets
Trans. Instr. Chem. Engrs., Vol. 32, 1954

The size and distribution of droplets in a 4 inch glass evaporator and pilot
plant were determined in this experiment. The experiments found that 95%
of the droplets were below 20 microns, but because of their low mass, they
made up a smal1 fraction of the total entrainment.

27. Grabner
Operating Conditions of the Down Cyclones for Steam/Water
Separation (In German)
Brennst. - Warme, Vol. 24, No. 5, 1972

This paper has not been reviewed except for the figures. The author gives
sketches of the downflow AEG separator and its performance curves. See
Appendix G, Number 9.

A-17
Vigorous Liquid
Entry
LEVEL OF LIQUID ENTERING TO POOL
MM
LINE,

/ Gentle Liquid
A Entry
ABOVE CENTER

CRITICAL LIQUID DOWNWARD VELOCITY


BASED UPON EMPTY DRAIN CROSS
SECTION AT CENTER LINE M/S

FIGURE A-8. CRITICAL LIQUID DOWNFLOW VELOCITY


FOR 5 VOLUME PER CENT CARRYUNDER

A-18
28. Gun
Wetness Fraction of Steam

Gunn, in this paper, recommends three methods for measuring the wetness of
steam. He said, in taking such a measurement, one might make the mistake of
visualizing the moisture to be in a cloud of droplets, more or less uniformly
distributed across the section of the steam main. Based on this picture, one
would attempt to sample the mixture by means of a perforated sampling probe.

Under practical conditions, the flow is turbulent, and there is a film of


water on the wall with the steam in the center. The three methods he re­
commends to sample the mixture are:

1. Use an apparatus to accelerate the steam flow to a critical velo­


city were the water droplets disperse. A sampling probe is then
used.

2. Use a steam meter followed by a steam separator.

3. Use a steam meter followed by several successive drains. Measure


the effluent from each drain.

29. Hitachi
Separator for Nuclear Power Plant
Japanese Patent 7366-32

This separator (Figure A-9) is supposedly better at lower volumetric flow


rates than the axial flow separator commonly used in BWR's. In this design,
steam/water is accelerated going down, while with an axial flow separator,
steam/water is accelerated as it goes upward through the vanes.

This separator consists of three concentric cylinders. From the reactor


vessel core, the steam/water enters upward through the outer annulus. At
the top of the outer annulus the steam/water makes a 180° turn and enters
vanes at the top of the intermediate annulus. The separated water is
channelled downward through the intermediate annulus into a manifold which
returns it to the circulation path. The separated steam flows downward on
the inside wall of the intermediate annulus before making a 180° turn and
exiting via a center pipe.

30. Hisana and Uenishi


Corrugated Plate-Type Steam Separator for Nuclear Power Generation Turbine
Karyoku Hatsuden, Vol. 22, No. 6 June 1971

Only the figures, tables, and plots in this paper were translated. A com­
parison is made of the corrugated and hooked vane chevrons. The comparison

A-19
Steam

Central Pipe

Vanes

Intermediate -
Annulus
Outer Annulus

Steam-Water

Water

Manifold

Figure A-9. Separator for Nuclear Power Plant

A-20
indicated that the corrugated chevron was more suitable for low pressures
of 14.2 to 42.6 psia at velocities of 32.8 to 65.6 ft/s and the hooked vane
chevron more suitable for higher pressures of 142 to 213 psia at velocities
of 6.6 to 13.1 ft/s. Also the authors have included performance curves and
photographs of bubble size and distribution at various locations on a turbine
blade.

31. Improvement in Vapor Separation From Circulating Liquid


British Patent 1153718

The apparatus (Figure A-10) has a two step separation process. The larger
bubbles are allowed to escape from the surface. The smaller bubbles are
separated by a cyclone and escape through an exit tube.

32. Japan Atomic Energy Research Institute


Experimental Studies of Steam-Water Separator for Boiling Water Reactors
NSJ-TR-62 (1966)

The experimental separator studies described in this paper were conducted in


three parts: Theoretical analyses, preliminary tests, and high pressure
tests. In the theoretical analyses, a force balance was conducted on a single
bubble using Feeble's empirical formulas for the drag coefficient, and the
equations were solved by numerical iteration. The analysis gave a qua!itative
estimation of the separation operation. In the preliminary tests, combina­
tions of three types of separators were tested in an air/water test loop
under atmospheric conditions. A separator with the most desirable vane
characteristics from the preliminary tests was scaled up and tested in the
high pressure steam/water loop at 853 psig. Lithium salt using a flame
photometer was used to measure carryover. In addition to measuring the separa­
tor efficiency (converging length, steam quality, pressure drop, and carry-
under) the flow stability was also obtained.

33. J. Karpisek
Centrifugal Bui 1t-in Water Separators for Heavy Duty Boiler Drums
Strojirenstvi, Vol. 18, No. 9, 1968

Three different kinds of centrifugal separators for heavy duty boiler drums
were tested: downflow (similar to Figure G-20), turbo (similar to Figure
G-19), and horizontal (Figure A-ll). Three types of vanes were tested in
the downflow design: semi-curved, fully curved, and radial. Sketches of
the separators and performance curves are included in this paper. Karpisek's
conclusion from this study was that the dov/nflow separator was adequate for

A-21
OUTLET

STEAM EXIT TUBE

CYCLOHE
HEATER

DOWNCOMER

RISER

MO o o

Figure A-10. Separation Apparatus.

A-22
Figure A-ll. Horizontal Separator.

A-23
his application.

34. Kiselev, Noskov, and Romankov


Investigation of Entrainment in Columns with Cyclone Contract Elements
Zhurnal Prikladnoi Khimii, Vol. 42, No. 9, 1969

The effect of the design dimensions on the performance of an axial flow


separator (Figure A-12) was determined in these air/water tests. For a
1iquid flow rate of 7.2 ft^/hr and air velocity of 72.8 to 98.1 ft/sec, the
following optimal dimensions were selected:

1. No. of blades - 6

2. Diameter - 3.5"

3. Separation height - 7.1"

4. Extraction region - 3.9"

The authors found that liquid entrainment increased with decreasing diameter
and increasing air velocity and water flow rate.

35. Kozlov and Eskin


On Calculating the Critical Loading of Separators
Teploenergetika, Vol. 11, No. 2, 1964

An evaluation of six different equations for the critical velocities of


baffle separators were made in this paper. Three parameters were re­
commended to take into account the influence of pressure, inlet quality,
and geometry on the correlation of experimental data for baffle separators.
The influence of pressure should be taken into account by,

q(s' - $")
(«")2

Where

a = surface tension
6' = spec. grav. of water at saturation condition
6" = spec. grav. of steam at saturation condition

The parameter indicating the influence of moisture is (yr^) where x =


quality.

A-24
Figure A-12. Axial Flow Separator with Castellations.

A-25
The formula for the critical velocity takes the following general form;

0.27
r(<S' - 6") F(d)
cr (l^xl
(6")2

Where

A = coefficient
n = exponent
d = diameterof baffle
W = critical velocity
cr

36. Kudirka, Burley and Moen


Flow Distribution Apparatus
U.S. Patent 3720046

Most dryers have low flow resistance which results in the steam flow passing
through in a non-horizontal direction. This tends to make the dryer less
effective and limits the height. In the concept described by this patent,
the passage area for the flow entering the distribution apparatus (Figure
A-13) from the bottom to the top and conversely for the flow exiting. The
open area into the dryer gradually increases upward along the perforated
plate and decreases downward in the other perforated plate as the flow leaves.

37. Kudira, Robbins and Moen


Gas-Liquid Separation on a Vane of a Radial Vane Separator
ASME Paper 66-WA/NE-ll

This paper describes the analysis and air/water experiments that were per­
formed to evaluate and confirm the performance of vanes to be used in a radial
vane separator.

For the 4 inche radius with 130° arc, the minimum isolated bubble that could
be separated can be represented by a general expression.

n+2 1 -n
t 1-n 1-n 1-n
D K R yl
R 9
Pi0'

A-26
0000000

oooooo

OQGQO

0 0 0 0

GOO

I
o o

Figure A-13. Flow Distribution Apparatus.

A-27
The authors use the Taylor stability criterion to determine average bubble
size:

ct9c

Y (Pi-g) (VVR)

= Bubble diameter
t = Water thickness of flow on vane
R = Vane radius
9 = Arc angle
n = Exponent according Re No. flow region
K = Constant
u-j = Viscosity of 1 iquid
p-j, = Density of liquid
a = Surface tension
V = Homogeneous nozzle velocity

The experimental results indicate that the bubble distribution at the entrance
and exit of the radial vane are similar. The separation analysis predicts
larger bubbles would separate first, leaving extremely small bubbles in the
stream near the vane exit. The authors explain this difference as being
due to the velocity distribution not being equal across the width of the
stream and/or the bubbles coalescing and breaking up continuously along the
vane.

38. Lang and Sonnenmoser


Moisture Separator/Reheater for Saturated-Steam Tubines
Brown Boveri Review, Vol. 63, No. 1, 1976

In Brown Boveri's first moisture separator reheater, they turned away from
the employment of wire mesh for separation purposes. Due to this and a
change in cost relationships, a modification of the Brown Boveri concept was
carried out. Instead of wire mesh grids, plate separators are used. These
plates effect a sudden change in direction of the steam flow and the water
particles contained are thereby separated out on the walls of the plate.
The water droplets are then suitably led away from the walls.

A-28
39. Lemeyis, Bushey, and Rabas
Progress in Moisture Separator - Reheaters for Nuclear Power Plants
Proceedings of the American Power Conference, Vol. 31, 1969

Woven wire mesh chevron separators show certain similarities in their


characteristics.
1. For a wide range of conditions, the moisture removal efficiency is
dependent on depth and mechanical design of the separator (wire
size, chevron spacing, etc.).

2. A rapid breakdown in the moisture removal efficiency occurs at some


characteristic vapor dynamic head. Extra depth will not alter the
breakdown point. The magnitude of dynamic head or flood point is a
function of the separator mechanical design and initial vapor qua­
lity.

3. The pressure drop through the mesh or chevrons is negligible. Dis­


tribution may be a problem.

Two factors should be considered in selecting a separation device to minimize


carryover.
1. Mesh must be mounted in a horizontal position or tilted no more than 15
degrees. Chevrons must be mounted within about 30 degrees of the vertical.

2. The allowable average design velocity for chevrons is about twice that for
mesh.

40. Marshal 1
Air-Water Separation Test Program
CEND-168
Air-water tests were conducted on a ful1-circle arc separator and tangential
arm separator. At the critical 1imiting flow rate, the annulus of the ful 1 -
circle arc separator completely filled with water and forced air down into
the downcomer. The test results indicated that separation occurs in the
downcomer (ful1-circle arc separator). The tangential arm separator was
recommended for further development.

41. Marshall
Carryunder in Gravity Separation of Air-Water Mixtures
ASME Paper 64-WA/HT-38
Air/water tests were conducted to investigate the effect of carryunder in
gravity-separation systems. For a Froude number (Vg/Dv)^ greater than
0.35, a double steam/water interface condition occurred and carryunder in­
creased substantially. Figure A-14 shows this double interface condition.

A-29
Vp = superficial downcomer velocity
g = acceleration of gravity
Dy = vessel diameter

A-30
Steam

Upper Interface

Separation Plenum

Vessel
Downcomer

Riser

Lower Interface

Steam/Water

Figure A-14. Gravity separation test section.

A-31
42. Marshall
Centrifugal Liquid Vane Separation
U.S. Patent 3507099 and British Patent 1038563

The steam/water mixture enters through the central tube and is redirected by
the taper to the curved arm vanes (Figure A-15). The taper at the upper end
of the central tube distributes the flow equally to the curved arm vanes.
The curved arm vanes create the centrifugal force and discharge the water
essentially tangentially on the inner surface of the outer jacket. The use
of the discrete curved arm vanes allow the steam to flow upward thereby
avoiding increased turbulence and pressure drop. The water exits downward
in the annulus between the central tube and outer jacket and then exits over
the water seal. The water seal is provided to minimize steam entrainment.

43. Marshall
Development of Steam Separation Devices
CEND-127, April 1961

This report delineates tests in which the disengaging height was measured.
A test drum, as shown in Figure A-16, was used for the tests. For some high
flow rates and submergence heights, the disengaging height was above the
highest probe (Figure A-16). Figures A-17, A-18, A-19 and A-20 illustrate
how steam flow rate, water flow rate, submergence and pressure affect the
disengaging height. A 12.5 inch riser was used for the result in Figure
A-19. This work can only be used to give an approximation of the disengaging
height.

44. Matanabe and Miyosho


Moisture-Separator/Reheater Unit for Nuclear Power Plant
Sangyo KiKai No. 293, February 1979

The separator element used in the moisture separator and reheater is of the
chevron type. Perforated plates are placed in front of the moisture separator
to obtain uniform axial flow. The wet steam flows zig zag across the chevron.
The condensed water drains vertically down. Performance curves for this
separator (dryer) are given.

45. Micheller and Patterson


Water-Steam Separation
U.S. Patent 3735569

This concept (Figure A-21) is a perforated jacket separator with a centrifugal


spinner similar to Design 5 of Appendix B. This separator is installed in a
cell to prevent interaction with other separators. The cel 1 should be square
to maximize the flow area, thus minimizing carryunder.

A-32
I

'/// /f;Z777777%>

Water/Steam

Figure A-15. Curved Arm Separator.

A-33
steaw

STEAM DRUM

-rNO.?
12"
4N0.4
12"
4 N0.3> SAMPLES
12"
120“

_
A-34

FROM
BOILER

TEMP.-AAA-•

ORIFICE

MIXING
CHAMBER

t
Figure A-16. Test Drum Internals-High Pressure Tests.
< P = 750 psia, Ws = 20000 #/HR, Ww = 169000 #/HR
■ P = 750 psia, Ws = 20000 #/HR, Ww = 169000 #/HR

❖ P = 1000 psia, Ws = 15000 #/HR, Ww = 169000 #/HR


V P = 1250 psia, Ws = 15000 #/HR, Ww = 169000 #/HR
A p = 1250 psia, Ws = 10000 #/HR, Ww = 169000 #/HR
iHT (REFERENCE WATER LEVEL TO STEAM-WATER INTERFACE)-INC

20

10

-12 -8 -4 0 4 8 12

SUBMERGENCE-INCHES

Figure A-17. Effect of Submergence on Separation Height.

A-35
O P = 750 psia, S = 9 inches, Ww = 171000 #/HR

O P = 1000 psia, S = 9 inches, Ww = 171000 #/HR

UJ

<

UJ
INT !
iHT (REFERENCE WATER LEVEL TO STEAM-WATER

STEAM FLOW #/HR x 10"3

Figure A-18. Effect of Steam Flow on Separation Height.

A-36
0 inches, Ws = 17000 #/HR
0 inches, Ws = 10000 #/HR
0 inches, Ws = 5000 #/HR
HEIGHT (REFERENCE WATER LEVEL TO STEAM-WATER INTERFACE)-INC!

WATER FLOW, #/HR x 10'3

Figure A-19. Effect of Water Flow on Separation Height.

A-37
□ S = 6 inches, Ws 25000 #/HR, Ww 172000 #/HR
O S = 9 inches, Ws 15000 #/HR, Ww 172000 #/HR
80 O S = 6 inches, Ws 20000 #/HR, Ww = 172000 #/HR
A S = 9 inches, Ws 10000 #/HR, Ww = 172000 #/HR

70

60

50

40

30

20

10

0 ___________________ (___________________ I------------------ -----------—I-------------------- ------------ L------- ----------------------l- - • ------------- -- I

700 800 900 1000 1100 1200 1300

PRESSURE - psia

Figure A-20. Effect of Pressure on Separation.

A-38
Figure A-21. Water steam separator.

A-39
46. Moen, Kudirka, Fitzsinmons and Burley
Advances in Boiling Water Reactor Steam Separation Systems
ASME Paper 69-WA/NE-S

This paper describes the evolution of the General Electric Separator. A


general description of the separator and dryer operation is included.

The following modifications were made to the chevron dryer to increase


its capacity.

1. Slanting chevrons

2. Increasing inlet area at the expense of exit area

3. Having variable inlet and outlet area

The tests using multiple separators compared to a single separator showed


no distinctive differences in performance.

47. Moen
Gas-Liquid Vortex Separator
U.S. Patent 3920876

This patent describes an axial flow separator with multiple skimmers aligned
one above the other. See Appendix B, Number 3. The multiple skimmers allow
for a decreased separator diameter, and thus a higher loading per area. The
restrictor in the first downcomer annulus throttles the water to the skimmers
The restrictor clearance was also chosen to minimize carryunder.

48. Moen
Nuclear Superheat Project Internal Steam Separation Development of Radial
Vane Steam Separators
GEAP 4272

Air/water tests were conducted for the following four separators:

1. Turbo (Figure G-19)

2. Archimedes Spiral (Figure A-22)

3. Tangential (Figure A-22)

4. Radial Vane (Figure G-14)

The radial vane separator was chosen for further testing because of its lower
pressure drop and larger volumetric flow capacity. The steam/water tests in-
3
dicated that the radial vane separator had a capacity of 5500 ft /hr of steam

A-40
Tangential

Steam

Water

Water Steam

Archimedes Spiral

Water

Water Steam

Figure A-22. Tangential and Archimedes Spiral Separators.

A-41
49. Moen
Nuclear Superheat Project Internal Steam Separation Development of Radial
Vane Steam Separators
GEAP 4272

This report pertains to the development of a full circle, radial vane


separator. See Appendix G, Number 6. Both experimental and analytical
information are included. The radial vane separator utilizes the centrifugal
force imposed on a steam flowing on a concave surface to bring about separa­
tion of steam and water. The steam which is separated flows off the ends of
the vanes.

50. Mozharov
Maximum Permissible Rate of Steam Flow Through a Separator
Teploengetika, Vol. 8, No. 4, 1961

The critical velocity for stripping the water film from a baffle separator
was derived in this paper. First, the formulas for the critical velocity
as derived by previous researchers were analyzed. Second, with the aid of
similitude theory characteristic parameters for the critical velocity were
determined. Finally, by processing experimental data using the characteris­
tic parameters, the following equation for critical velocity was recommended:

rkp . 200 (0B)-1/3

Where

W"kp = Critical velocity (m/s)


G^ = Moisture loading on wetted perimeter (kg/ms)
a = Surface tension (kg/m)
3
y" = Specific gravity of steam (kg/m ).

51. Panasenko and Koslov


Investigation of Separating Equipment for High Capacity Drum Boilers
Teploenergetika, No. 8, August 1962

High pressure steam/water tests were conducted by the authors for the
vertical baffle separator. The critical velocity was determined to be
a function of system pressure and inlet quality.

A-42
U = (1.4 x 1.8(1-X))10"-0056 p

and

Where U(meters/second)
p(bars)

The vertical baffle was 5.9 inch high, 3.2 inch wide, .4 inch bending radius,
.2 inch pitch, and .02 inch thick. The gap between plates was 0.18 inches.
The width of the secton was 4.7 inches.

The reasons were given for spread in the high pressure steam/water data:

1. Instability of the separated moisture film

2. Uneven generation of droplets.

52. Patterson, Micheller and Carney


Centrifugal Separator Having Perforated Can Wall
U.S. Patent 3614863

This patent describes a nuclear separator used by Combustion Engineering.


See Appendix B, Number 5. The water, after being separated by the vanes,
discharges through the perforations, and the steam exhausts upward through
the core. Major advantages of this separator are its compact size and high
moisture extractions at low flow rates.

53. Peter
The Water Separator of the Sulzer Monotube Steam Generator
Sulzer Technical Review, March 1969

This separator is relatively tall with a tangential entry. The water level
is maintained fairly constant. The water exits downward and the steam vents
through at the .top of the separator. This separator removes 95-99% of the
moisture from the inlet steam.

54. Petrick
A Study of Carryunder Performance in Vapor Liquid Separation
A.I. Ch.E Journal, Vol. 9, No. 2, 1973

An analytical and experimental study of the carryunder problem was conducted


by the author to establish the effects of certain parameters. Figure A-23 is
a schematic of the system that was studied. An analytical study yielded the
characteristic parameters that affect carryunder. The air/water and high
pressure steam/water data was corrected for XD/XR in the range of 3 to 64.

A-43
Steam Out

Separation Plenum

Downcomer

Riser

Figure A-23. Circulation Loop.

A-44
In the range of 0.1 to 3,

#=-0.6 Log
Vq,l
Ve (^1/2(VT ^F)
XR

The author indicated that these relations should be prudently used when
applied to large systems. He recommended the following equation for down­
comer slip ratio:

0.4 0.2
!a 0.63 x fl
1-x pg
VL

= Downcomer quality
= Riser qualtiy
V 1 = Velocity of gas phase in riser
99
Ve = Velocity of entrained steam
a = Surface tension
y = Viscosity of liquid
= Density of 1 iquid
Pg = Density of gas
D = Riser diameter
H = Interface height
G = Liquid mass velocity

A-45
55. Petrick
A Study of Vapor Carryunder and Associated Problems
ANL 6581

Carryunder in a circulating loop similar to a PWR or BWR was studied. In the


study test, two phase mixture flowed upwards through a riser. At the top of
the riser near the interface, the liquid made a 180° turn into the downcomer
and the gas continued to rise and left the circulating loop. An extensive
theoretical analysis of the carryunder phenomena in the downcomer was made
followed by a dimensional analysis that yielded dimensionless groupings which
were used in correlating the data. Air/water and high pressure steam/water
tests for carryunder were conducted, and the following empirical correlation
was determined.

64

Where

Xp = Quality in downcomer
XR = Quality in riser
Vg = Actual gas velocity
Vp = Downcomer gas velocity
a = Surface tension
y = Liquid viscosity
p-j = Liquid density
pg = Gas density
D = Diameter
H = Interface height
G = Liquid mass velocity in riser

56. Moisture Separator-Reheaters


Heat Engineering, January-March 1976

A stainless steel hooked-vane chevron was chosen for the moisture separator-
reheater described in this paper. The design point is approximately 25 per­
cent below the maximum permissible to assure a conservative margin of no
more than 7.5 percent maldistribution.

A-46
57. Place
Combustion Engineering
Combustion Engineering, Inc., 1966, Chapter 8

This chapter is a review of steam and water separation technology, measuring


techniques, primary and secondary separators, and baffling.

58. Prins
Aspects of Two Phase Gas-Liquid Separation Related to Nuclear Supply Systems
Thesis Del ft U. of Tech., Netherlands, 1971

Prins conducted an extensive literature survey in working toward his three


objectives which were:

1. An experimental study of gravity separation.

2. An experimental study of an axial upcomer separator with a con­


verging-diverging vortex tube.

3. The derivation of a two-dimensional model to describe the separa­


tion process.

The model for the gravity separation consisted of a water and steam mixture
flowing upward through a riser, and at the interface level, the steam continued
flowing up and out through an exit nozzle while the water circulated through
the dowmcomer. Prins conducted carryunder tests and compared his results with
the results of several investigators. He concluded that water velocity turned
out to constitute the main influence on carryunder, and that the threshold
velocity varied between 0.72 to 0.92 ft/sec.

In his analysis of the upflow separator, (See Appendix G, Number 10), the
author listed three potential problems in the design:

1. Inlet region (vanes)

2. Vortex tube (separtion length)

3. Outlet region (skimmer or vortex finder)

In his converging - diverging separator, the author used eight 45° flat plate
blades and a converging section inclined at about 14.5°. A second tube was
installed in the vortex finder to remove the gas core and allow the water to
pass. The author recommended using carryunder and carryover efficiency to
evaluate the performance of separators. He also discussed a compound effi-
cieny;

A-47
nc " ^cu
Ws

Where

W$ = Steam flow rate


WCu = Steam carryunderflow rate
Ww = Water flow rate
Wco = Water carryoverflowrate

The tests of the converging-diverging separator displayed some basic problems


of the vortex finder. They were:

1. Hydraulic jump

2. Remixing phenomena of separated phases after being diverted by the


vortex finder.

The overal1 pressure drop was adequate when compared with other separators.

A comprehensive analytical evaluation was made of the separation process in


the converging-diverging separator. The method consisted of using two-dimen­
sional mass and momentum equations of the two-phase flow separation and
applying a N-Strip method. The solution had numerical instability.

59. Ravese and Bentley


Vapor-from-Liquid Separator Apparatus
U.S. Patent 2648397

The separator described in this patent is used in fossi1 boilers. (See


Appendix G, Number No. 7). The turbo vanes create the centrifugal force to
separate the steam and water. The water is skimmed off the inner tube and
returns down the annul us. The moisture laden steam exits through the center
core and is dryed by the corrugated chevrons.

The advantage of this separator is its compactness and good carryover perfor­
mance for the tested flow conditions. For the following conditions, this
separator has negligible carryover:

A-48
Pressure = 1250 psi

Steam flow rate = 14236 Lbm/hr

Total flow rate = 63077 Lbm/hr

Water level = center of separator

The separator pressure drop is 1.3 psi.

The pickoff ring (collector sleeve) is sized for optimum performance.

Collector Sleeve Area


Performance
Cross sectional area of primary tube

82 to 90% Best

52 to 82% Improves

90 to 90% Progressively decreased

Above 95% Sharply decreased

The author's experiments showed that the "single-stack" corrugated plate


arrangement was less desirable than the two-stack arrangement.

60. Ravese
The Application and Development of the Turbo Steam Separator
Combustion Engineering, July 1954

The design and performance of a turbo separator is given in this paper. The
separator described is used in fossil circulation boilers. See the discussion
of the turbo separator in Appendix G, Number 7.

61. Regehr
Theory of Water Separation
Von Karman Institute for Fluid Dynamics
Lecture Series 70, May 1974

Regehr used a theoretical approach in analyzing the operation of three water


separators: centrifugal, fiber packing, and impingement vane. He gave several
interesting pieces of information about the centrifugal separator.

1. The tangential velocity increases with increasing distance from


the chamber wall according to Wu x rn = constant. In the core,
n is close to -1, and at the periphery n approaches +1.

A-49
2. Reentrainment of drops from the wall occur when Pw^> 1250 Lbm/sec^
ft.

3. The drop size at which the fractional efficiency is 50% can be


determined from the following equation:

. (18Nq wri rt)


(pL-Pg)Wui2

Where

ng = dynamic viscosity of gas

wri = radial velocity at ri

wui = maximum tangential velocity


ri = radius of cyclone discharge

Pg = gas density

PL = liquid density

Summarized as follows:
8
1. The efficiency of wire mesh is Ep 1 - exp (-
301 wz)

Where

P = voidage

dp = fiber diameter
W = fractional efficiency

Z = depth

and

d
I
¥3 + fi'1'2 + f2'i + f3 d
F

fii = f(Reynolds number)


dy = drop size
CApLdT2
¥ (Stokes parameters) =
18ngdy

A-50
Where

C/\ = face velocity

P[_ = liquid density

ng = dynamic viscosity of gas

2. The maximum face velocity of wire mesh is given by the Souders-


Brown expression

^Amax= = 0.107

3. For wire mesh 3.9" thick and 97.5% free volume, Nonebel recommends
that thepgC^ should not exceed 11.2.

4. For higher pressures, values of CA are lower in the range of 3.28


ft/sec.

Some of the interesting facts regarding the impingement vane separator


are:

1. For air/water at normal conditions, the maximum face velocity for


vertical flow is 9.8 to 19.7 ft/s, and for horizontal flow it is
approximately 65.6 ft/s.

2. The resistance factor for vane separator is a function of Reynolds


number. The efficiency is a function of drop size.

3. Figure A-24 shows the performance of a Euroform hooked-vane separa­


tor and Figure A-25 shows the Euroform vane elements.

62. Robbins
Analyses of the Radial Separator
GEAP-4012

This paper contains a mathematical analysis of the performance of the radial


vane steam separator. The analysis indicated that if the Reynolds number is
high enough such that the drag coefficient is constant, bubble diameter is
not affected by velocity or vane radius.

63. Robbins
Internal Steam Separation Shown Successful in German Nuclear Plants
Electric World, July 24, 1967

Robbins discussed GE's technological advancement in internal steam separation


at the KRB plant. The 69 separators used were 13.25" diameter, 56" long and

A-51
pCAl (kg/ms2)

0 20 100 200 400 700


10 Face Velocity (steam)in/s
%
WETNESS

FACE VELOCITY (AIR) H/S

Figure A-24. Wetness of 2 Euroform Water Separators.

Figure A-25. Vane Elements.

A-52
mounted on 6" diameter standpipes. The Peerless hook vane type dryers com­
plemented the separator.

64. Robbins
Steam Separation Reduces BWR Size
Electrical World, March 25, 1963

The author describes the steam separation equipment in a BWR. He indicates


that in reasonably sized plenums, the flow is distributed uniformly to the
separators. The water level did not vary substantially. Two types of test
facilities were used in developing and choosing the separator dryer systems;
air/water at atmospheric conditions and high pressure steam/water.

65. Rouhani
Separator for Mixture of Steam and Water
Swedish Patent 7309949-1

This patent describes an axial upflow separator. For a description of the


vanes see Appendix E, Part 1. The steam in the separator cone exits through
a central pipe. The water f1ows through a perforated conical plate into the
separator downcomer.

66. Sandstrom and Oddving


Development of Steam-Water Separation Systems for ASEA-ATOM
Boiling Water Reactors
Reactor Systems and Components

The steam/water separation described in this paper is based on a two-step


separation process. Centrifugal force is used in the first step and Peerless
steam dryers used in the second step. Air/water and high pressure steam/water
were used in the development of this separator. The following are some re­
sults from the separator tests:

1. The upcomer type separator gave lower pressure drop than the down­
comer type.

2. Twelve vanes in the inlet geometry gave the best results.

3. The smaller number of holes in the perforated outlet cone decreased


carryunder. The flow area is a function of water flow rate.

4. Separator lengths above 3.3 feet did not significantly improve the
performance.

5. To reduce pressure drop, efforts were made to improve the inlet


geometry.

6. A pre-dryer reduced water content in the steam leaving the separa­


tor .

A-53
67. Seay and Lowry
Literature Survey of Gravity Separator
CE Report, December 1976

This 1iterature survey indicates that most of the pertinent literature re­
lated to gravity separation is for low steam release rates and low pressures
The authors make a calculation to determine if gravity separation occurs in
a steam generator. (For flow wherein liquid droplets are dispersed in a con
tinuous gas phase, a critical Weber Number of 13 was used to calculate the
maximum droplet radius.) The force balanced described by the following
equations indicated gravity separation occurred in some areas.

2 rmax Pc (Vq ~ V&) ^


^ecrit a

24 0.687 0.42
cD (1 + 0.15 Rep 5
Rep
1 + (4.25xl0“4)(Re -1,16)
r

and

Rep = DpPc |Vg - V*|

yc

Where

Wecrit = critical Weber number


Pc = density of continuous phase
Vmax = particle radius
Vg = gas phase velocity
V«, = liquid phase velocity
a = surface tension
Rep = particle Reynolds numbers
Dp = particle diameter

yc = absolute viscosity of continuous phase


Cp = drag coefficient

A-5 4
68. Simon
Steam-Water Separation in Boiling Water Reactors
C. Theoretical and Experimental Determination of the Operating
Characteristics of Cyclones for Steam Water Separation
Konstruction, Vol. 23, No. 11, 1971

Steam/water separation by means of an AEG downdraft separator (See Figure G-22)


was investigated from both a theoretical and experimental basis. In the
downdraft cyclone, a steam/water mixture enters upward through a central
transport tube. The flow is the divided into four 180° turning channels and
travels downward through an annular set of vanes where separation occurs.
A vortex is created as a result of the separated steam turning and flowing
up through a central exit tube. As steam flow increases, water carryover
decreases, until the steam vortex drops below the aperture. An aperture
(restriction) is provided at the water outlet apparently to maintain a
sufficient steam vortex and/or to minimize re-entrainment.

The author indicated that perfect separation cannot be achieved. He contends


that there wi11 always be a certain miminum amount of water on the entire
phase boundary of the steam vortex. There will be water droplets of a smal 1
enough diameter that cannot be distinguished from the steam. A part of the
steam will be entrained in the water as fine steam bubbles.

A pre-dryer is incorporated at the exit of the steam pipe. Some residual


centrifugal force is left in the steam when it enters the pre-dryer. The
heavier water droplets are tossed on the wall and form a water film. The
water film is peeled off by a skimmer. The author indicates that some of
the moisture falls out in the space between the separator and dryer. At
1000 psi this occurs below a steam velocity of 2.3 to 3.3 ft/sec.

69. Sokolowksi
Formed Vane Elements for Liquid-Vapor Separators
U.S. Patent 3805496

Vane elements described in this patent have a zig-zag cross section with
hooks open to the flow (See Appendix C, Figure C-6). The moisture collects
in the hooks, drains down to a trough, and the water escapes through drain
pipes. Each vane can be fabricated from a single sheet of metal formed by
stamping, bending and/or roll forming in the desired shape. This reduces
construction costs and a 11ows for close tolerances. By incorporating per­
forated plates at the entrance and exit of the vane elements, flow distribution
through the vanes improve.

A-55
70. Sokolowski
Vertical Steam Drum
U.S. Patent 3751886 abd French Patent 2275731

Two stages of separation are uti 1 ized to remove the moisture. The first
stage is a separator with turbo vanes and a downcomer to channel the separated
water (See Appendix B, Number 2). The exit orifice through which the steam
escapes is sized to minimize carryover and probably carryunder. The second
stage of separation consists of two tiers of hooked vane dryers. A unique
feature of this dryer is that each chevron is fabricated in one piece, thus
avoiding a tedious welding process. A perforated plate is installed at the
dryer exit to improve flow distribution.

71. Sorokin, Demidova, and Kuz'min


Principles of Drop Separation from Vapor or Gas Streams
Chemical and Petroleum Engineering, No. 7-8, 1968

Separator (dryer) designs (Figure A-26) as fol1ows were compared with re­
spect to their ability to remove moisture:

1. waveform horizontal baffles

2. zig zag horizontal baffles

3. raschig rings

4. wire mesh

5. angle

6. vertical baffles

7. inverted channel

The performance data for these separators were described in terms of a


"Kutateladze" number. The authors' conclusion was that vertical and
inclined baffles separators were the best performers. They could operate
at high velocities and supply lower moisture steam than the others evaluated.

72. Stern
Internal Steam Separation, Pumping and Flow Path in Boiling Water Reactors
U.S. Patent 3342690

The separator (Figure A-27) described in this patent is designed for a boiling
water reactor. The separator is.gear drive mounted so that the steam/water
mixture from a boiling water reactor core passes through the separator which
is connected to a circulating pump. The separator provides for circulation
and separates the steam and water.

A-56
Waveform Zig Zag Rashig Wire
Horizontal Horizontal Rings Mesh
Baffles Baffles

Angle Vertical Invertical


Baffles Channel

Figure A-26. Separation Devices.

.A-5 7
Steam Dome

Upper Bearing

Arc Separator

Lower Bearing
Dri ve
Gear

Downcomer

Circulating
Pump

Figure A-27. Arc Separator.

A-58
73. Stern
Improvements In and Relating to Vapor-Liquid Separating Apparatus
British Patent 3342690

The separator described in this patent is a centrifugal separator of an bottom


inlet type (See Appendix G, Number I). The liquid/vapor flow enters upward
through an inlet conduit before being separated by the vanes. The liquid
is slung against a converging barrel and flows downward in the annular gap
between the barrel and the inlet conduit. There is a second set of vanes at
the exit of the annular passage to remove excess steam carried under. The
water flows down the downcomer of the separator, and the steam escapes upward
through both the downcomer and center core.

74. Steam Separator in BWR


Japanese Patent 41-18365

The unique feature of this separator (Figure A-28) is that there are several
stages of vanes, each remvoing more of the steam from the water. Initially,
the steam/water enters the primary vanes where the majority of the steam is
separated and is allowed to flow through a center pipe to the steam dome.
Water with the unseparated steam flows through si its into secondary vanes
where the process is repeated. There is also a similar tertiary stage. See
Appendix E, Part 3.

75. Sterman
On the Theory of Steam Separation
Soviet Physics, Technical Physics, July 1958

Sterman investigated the differential equations involving carryover, and deter­


mined the characteristic parameters. He correlated carryover data for pressures
from 1.27 atm to 110 atm in terms of these parameters. The carryover for sub-
critical steam velocities was determined by using the following equations:

,9
CO = 6.1 x 10

A-59
0O 0 0

Steam
Water

Steam

Figure A-28. Axial Flow Separator with Secondary Vanes.

A-60.
Where

CO = carryover
vs = velocity of steam
H = height of steam space
g = acceleration due to gravity
v = kinematic viscosity of 1iquid
Y = specific weight of 1iquid
y" = specific weight of steam

76. Tischer
Studies on Cyclones for Steam Separation in Boiling Water Reactors
EURAEC-2029

This report has a detailed discussion of the vane design. One of the better
downflow separators (See Appendix G, Number 9) tested has 8 vanes that were
4.7 inches in length. The steam extraction pipe diameter and separation pipe
diameter for one design are 3.5 inches and 6.6 inches respectively. The
mixture outlet angle for the axial entry (See Figure G-21a) is as follows:

3 - ARC Tg

a = Vane outlet angle, measured in the arc plane with respect to


the cyclone axis

a = Eccentricity of the vane outlet edges with respect to the


cyclone axis

r = Radius

For the tangential flow entry (See Figure G-22b) with spiral vanes.

(3 = - ARC Tg -iL/J._JL >


4 r

Where

h = Pitch of the sprial

A-61
77. Tischer
Studies on Cyclones for Steam Separation in Boiling Water Reactors
EURAEC-2127

This study is one of the many EURAEC Reports. An objective of this study
was to conduct air/water separator tests that would be comparable to high
pressure steam/water experiments. Several modelling criteria were followed
in conducting these air/water tests. First, the same volumetric water and
air flows as in the reactor were used. Second, because the pressure drop
affects the separation process, and in order to maintain a comparable pressure
drop for the air/water tests, a throttling valve was placed in the steam
exit line. Third, in order to determine the influence of properties on the
bubble separation inside the cyclone, tests were made at 68 and 176°F. At
these temperatures, there were differences in the surface tension and kinematic
viscosity. The conclusion was that both air/water and steam/water tests show
the same tendency. For example, a cyclone separator which shows excessive
air carryunder in air/water tests, most probably does not function in steam/
water either.

78. Traube and Vollradt


Investigations on the Separation of Flowing Gas/Liquid Mixtures
by Centrifugal Forces
Chem-Ingr.-Tech., Vol. 34, 1962

Guide vanes for an axial cyclone were tested using air/water (See Appendix E,
Part 1). The authors found that the rotation producing the separation simul­
taneously reduces the size of the gas bubbles, which hindered separation.
When the rotation is increased, this hindering trend can increase to a greater
extent than the trend toward promotion of the separation. Consequently, for
every 1iquid flow characterized by velocity and gas content, there is an
optimum rotation, the exceeding of which results in a decrease in separation.

79. Weismantel, Razgaitis, and Buxton


Influence of Cone Design Upon Liquid Entrainment Separation for a Steam
Cyclone Separator
ASME Paper 76-WA/PWR-6

A downflow separator (Figure A-29) was tested in a steam/water loop at a


pressure of 150 psig. Five different cases were tested:

1. Short 30° cone

2. Short 20° cone

3. Long 30° cone

4. Short 20° cone

A-62
Inlet

Vortex Finder

Separator Shell

Separated Liquid
Outlet

Figure A-29. Test Separator.

A-63
5. No cone

Q
The short cone geometries were superior to the cone-less case. The short 30
cone exhibited the best performance and the long 20° cone the poorest perfor­
mance. This performance factor is defined as follows:

Where

P = pumping power
n = separator efficiency
c = With cone
o = Without cone

80. Wilson, Littleton, Yant and Meyer


Primary Separation of Steam from Water by Natural Circulation
ANCP 65002, April 15, 1965

Wilson performed water carryover tests for various conditions. For an


18 inch diameter channel, pressure from 300 to 1200 psia, and steam dome
heights from 1.75 to 7 feet, the test data was correlated using dimensionless
groups:

-.08 64 051 1.28

Where

and

Uc = Critical steam velocity


H = Separation height
a = Surface tension
g = Acceleration due to gravity

A-64
vg = Gas kinematic viscosity
pg = Gas density
p]_ = Liquid density
Ap = Density difference between phases

Some of the typical test result are shown in Figure A-40 for 1200 and 1000 psi.

81. Wilson, Grenda, Klumb, Littleton, Meyer and Yant


Primary and Secondary Separation of Steam from Water by Mechanical Means
ACNP-65003

The results of an extensive steam/water separation program are delineated


in this report. Four types of separators were analyzed and tested.

1. Downcomer (understand hydrodynamic principles), Figure A-30.

2. Upcomer (reduce carryunder), Figure A-31.

3. Horizontal (use with low head room), Figure A-32.

4. Hooded cross over (efficient separation), Figure A-33.

In the analysis of the separators, the separation length was calculated


based on the average bubble size and vortex geometry. Various separator
parameters such as vane design, separator diameter, inlet nozzle, and
outlet nozzle were optimized during these tests.

Additional tests of a wire mesh dryer, chevron dryer, and gravity separation
3
were conducted. The wire mesh was 6 inches thick with a 5 Lb/ft mesh in
layers. The chevron dryer was composed of hooked-vanes. The maximum steam
release rate at 1000 psia was approximately as follows:

Wire mesh = 45000 Lb/hrft^

Chevron dryer = 50000 Lb/hrft^

Gravity separation at 2ft = 19000 Lb/hrft2

Gravity separation at 4ft = 25000 Lb/hrft2

82. Wolf and Moen


Advances in Steam-Water Separators for Boiling Water Reactors
ASME Paper 73-WA/PWR-4

The development of the GE axial fl ow separator is described in this paper.


Sketches of the separators, a description of the operation, and plots of the
performance are included. See Appendix B and Appendix G.

A-65
STEAM

k
l
l
I

MIXTURE

WATER

Figure A-30. Downflow Separator.

A-66
Figure A-31. Upcomer Separator.

A-67
Figure A-32. Hooked Crossover.

A-68
Figure A-33. Horizontal Separator.

A-69
83. Wilson and Grenda
Removal of Entrained Moisture from Steam Using Natural Separation and
Mechanical Dryers
ACNP 6105, April 3, 1961

Three different dryers were tested: centrifugal (Figure A-34), chevron


(Figure A-36), and mesh (Figure A-38). The paper is not very clear as to the
design of the centrifugal dryer. The limiting factor of the centrifugal dryer
is the pressure drop across the dryer exceeding the hydrostatic head available
in the drain pipes. Figure A-35 contains performance curves for the dryer
at pressures of 300 and 600 psig. The inlet quality was approximately 1%
for all the dryer tests.

The chevron impingement dryer (Figure A-36), depends on inertia, centrifugal


force and coalescence. The performance is shown in Figure A-37.

The wire mesh dryer (Figure A-38) provides a lot of surface area for the
droplets to impinge upon. The mixture continually changes directions as it
flows through the mesh. The performance of a wire mesh (6 inches thick,
5 Lbm/ft^, and an angle 15° to the horizontal) is shown in Figure A-39. When
3
the mesh density was increased from 5 to 12 Lbm/ft and the thickness from
6 inches to 12 inches, the performance only improved by approximately 10%.

The mesh of rolled dryer construction had a 13 to 18% lower flow rate than
the mesh of the layer construction. No measured effect could be detected by
increasing the inlet angle from 15 to 20°'

Figure A-40 shows separation results for natural moisture separation. The
moisture measurements were made at a height of 24 inches from the liquid
surface.

84. Wilson, Grenda, and Patterson


Steam Volume Fraction in a Bubbling Two-Phase Mixture
ANS Trans., Vol. 4, No. 2, 1961

The authors conducted an experiment to determine the steam volume fraction


in the process of bubbling steam through saturated water. They recommend
the following equation for steam volume fraction.

A-7Q
V %
a
< 2 K = 0.68 a = 0.62
g A p

V 1 %

a >_ 2 K=0.88 a = 0.40


9 A p

a = Void fraction
K = Constant
g = Gas density
f = Liquid density
Ap = Density difference between phases
o = Surface tension
g = Acceleration due to gravity
V = Velocity of steam
d = Vessel diameter

85. Vollradt
Steam/Water Separators for Boiling Water Reactors
Kerntechnik, Vol. 12, No. 7, 1970

Various separation systems used in BWR's were reviewed. The author compared
air/water screen tests with steam/water verification tests. For an upflow
cyclone at KRB and downflow cyclone at KWL, he compared test and operating
performance.

The author emphasized that an important consideration of any separator design


was its testability. (Whether a reliable test can be conducted before its
installation into a reactor.)

In the air/water screen tests, the carryover appears to be measured by a


calorimeter and the carryunder by the aP across a valve. The air/water
facility had a throttling device upstream of the separator in the steam line.
The author said that by varying the choke, it was possible to compensate for
the lower density of air compared to steam for the effect of pressure drop.
The carryunder and carryover should be adjusted by varying the pressure drop
across this throttling device.

A-71
A-A

Figure A-34. Centrifugal dryer.

A-72
12

II

10

4 8 12 16 20 24 28 32 36
STEAM FLOW RATE LB^/HR X ICT

Figure A-35. Centrifugal Dryer Test Results.


POCKETS TO COLLECT MOISTURE

SPACER

MIXTURE
FLOW
>
l

TYPICAL VANE SPACING

Figure A-36. Chevron Type Dryer.


800 PSIG

700 PSIG

JU)
cc
D
f- 600 PSIG
C
5
i
b-
UJ
.J
t-
D
O
500 PSIG

300 PSIG

STEAM FLOW - LB /HR-FT 2X IQ3

Figure A-37. Operating Characteristics of Chevron Type Dryer.

A-75
LAYER TYPE OF CONSTRUCTION
STACKED TOGETHER

METAL BAND --------- \


EACH END TO
PREVENT FRAYING

STACK TO DESIRED HEIGHT ^$||||||

DESIRED
desired
HEiGHT
HEIGHT

(A)
LAYER TYPE OF
CONSTRUCTION
(B)
ROLLED TYPE OF
CONSTRUCTION

Figure A-38. Methods of Fabricating Mesh Type Dryer.

A-76
600 PSIG
MOISTURE
O U TLE T

400 PSIG

300 PSIG

O 5 10 15 20 25 30 35
STEAM FLOW - LBS/HR-FT2X I03

Figure A-39. Operating Characteristics of a 5-LB-PER-CU-FT,


6-Inch-Thick Mesh Type Dryer.

A-77
PERCENT
-
MOISTURE

1000 psi

STEAM RELEASE RATE (Lbm/hr Ft )

Figure A-40. Moisture Carryover Tests.

A-78
86. Yarden and Faust
Improved Moisture Separator Reheaters for Nuclear Steam Supply Systems
Power Engineering, January, 1976

The wet steam enters the vane type separators at 12 to 15% moisture (by
weight) in various-size droplets. For compactness, these vane separators are
inclined. The vane separator has replaced the wire mesh demister.

87. Yeh and Zuber


On the Problem of Liquid Entrainment
ANL-6244, October, 1960

Zuber and Yeh gave a very good description of the mechanisms governing liquid
entrainment. They indicate that conditions leading to bubble formation deter­
mined the droplet size. Droplets are formed by two mechanisms:

1. Bursting of bubble dome (on the order of a few microns).

2. Detaching of liquid jets which are formed by water filling in the


cavity left by the bursting bubble (on the order of 100 microns).

Bubbles larger than 1.27 inches in diameter produce droplets only by the
collapse of the bubble dome.

Zuber and Yeh discuss the mechanism for water entrainment in a vessel.
Figure A-41A shows a vessel with water level H^. As vapor is bubbled through
the water, drops are generated by bursting bubbles. For those large droplets
with a settling velocity larger than the vapor rise velocity, the drops will
rise to a height (Region 11 of Figure A-41B) before falling back to the 1iquid
surface. As the vapor flow rate is increased, the water level increases due
to the increased void fraction. In this case, some of the larger droplets
escape (Region II of Figure A-41C). A still further increase causes inter­
action between the bubbles and the formation of vapor siugs. As these vapor
slugs reach the surface 1iquid sheets are generated (Region II of Figure A-41D).
Zuber and Yeh presented the results of Sterman, Figure A-42. This figure
shows that liquid entrainment is directly proportional to constant C, and the
reduced vapor velocity to the n power, E = CUn. The power term increases
according to three cases shown in Figure A-41.

88. York
Performance of Wire-Mesh Demisters
Chemical Engineering Progress, Vol. 50, No. 8, August, 1954

A wire mesh fabricated with a wire diameter of .011 inches, wire mesh density
3 2 3
of approximately 12 Lbm/ft and surface area of 110 ft /ft has an efficiency

A-79
A - 80

Figure A-41. Mechanisms Governing Liquid Entrainment.


n •* 7 r 20

TRANSITION POINT(b)

n « 3+4

l I t I l l i i i i i i i i i i

0,000

Figure A-42. Entrainment as a Function of the Reduced (Superficial)


Vapor Velocity for Water at 185 ATM and for the Height
of the Vapor Dome 8=0.465 m.

A-81
of 99.9% over a wide range of vapor velocities.

High separation efficiencies were obtained by decreasing the wire diameter.


Section 2

LISTS

BAFFLE PLATE SEPARATORS

Most of the variations of the baffle plate separators were described in either
Appendix C or the first part of this Appendix. The following additional references
for the most part are patents and bulletins. These may be useful for an in-depth
study of dryers.

1. Alen'kin, Golub, Gol'dshtein, and Rozen


Efficiency of Steam Cleaning in Evaporator Apparatus with Louvre
Traps
Khim. Prom., Vol. 10, 1967

2. V and VR Separators
Bulletin 3210-1, Burgress Industries

3. Day
Air Washer and Li ke
U.S. Patent 1402147

4. Farr, Culbert, and Segundo


Mist Col lector
U.S. Patent 3254475

5. Germerdonk and Gunther


Efficiency of Drop Separator in Vertical Gas Stream (In German)
Chemie-Ing.-Techn., Vol. 41, No. 11, 1969

6. Golden
Vane-Type Separator
U.S. Patent 3517486

7. Good
Vapor-Liquid Separator Having Improved Vane Space Means
U.S. Patent 3520116

8. Halter and Chance


Vane Type Separator
U.S. Patent 3405511

9. Horton and Dixon


Apparatus for Separating Liquid from Liquid Entrained Gas
U.S. Patent 3490210

A-83
10. Kienbock and Kirn
Development of the Centrifugal Separators with Agglomerator for
Nuclear Power Industry (In German)
VGB Kraftwerkstechnik, Vol. 55, No. 8, 1975

11. High Efficiency Gas Separators for Removing Liquids from Gas
King Tool Bulletin

12. Vane Mist Eliminators


Bulletin No. 6, Peco Equipment Corporation, 1976

13. Peyrolorgue
Dryer-
French Patent 2234917

14. Liquid Separator


Steinmuller Bulletin

15. Sugimura
Impingement Separator for Gas-Liquid Mixtures
U.S. Patent 3938972

DOWNFLOW GAS - LIQUID SEPARATORS

Some of the early separator development work was for the downf1ow type. This type
separator has application in other areas besides nuclear such as the petroleum
i ndustry. Many EURAEC reports, which are too numerous to list, are on the develop­
ment of a downflow separator for a BWR. Some of the more important references are:

1. Fleeter and Ostrach


Liquid Gas Cyclone Separator
AFOSR 66-1400, June 1966

2. Poliak
Steam-Water Separation in Boiling Water Reactors
B. Design of Steam-Water Separators and Steam Dryers (In German)
Konstruktion, Vol. 23, No. 11, 1974

3. Poliak and Work


The Separation of Liquid from Vapor, Using Cyclones
Transactions of the ASME, January 1942

4. Lohr
Steam-Water Separation in Boiling Water Reactors
A. The Steam-Water Separation as Parts of the Cooling Circuit
(In German)
Konstruktion, Vol. 23, No. 11, 1971

DUST SEPARATORS

Even though the techniques in the design or analysis of a dust separator may not be
directly applicable to a steam-water separator, they are useful in gaining ideas

A-84
that may be incorporated into a design or to help one become aware of the equations
that might be needed in an analysis of steam water separators.

1. Alexander
Fundamentals of Cyclone Design and Operation
Australasian Institute of Mining and Metallurgy
No. 152-153, March-June 1949

2. Bloor and Ingham


On the Efficiency of the Industrial Cyclone
Trans. Instr. Chem. Engrs., Vol. 51, 1973

3. Boldyrev
Analysis of Solid Particle Movement Along the Generators of a
Hydrocyclone
Teoreticheskie Osnovy Khimicheskoi Teknologii, Vol. 8, No. 2,
March 1974

4. Lapple
Processes Use Many Collector Types
Chemical Engineering, May 1951

5. Leith and Mehta


Cyclone Performance and Design
Atmosphere Environment, Vol. 7, Pergamon Press, 1973

6. Leith and Licht


The Collection Efficiency of Cyclone Type Particle Collectors
AICHE Symposium Series, Vol. 68, No. 126, 1972

7. Muschelknautz and Krambrock


Aerodynamics Veiwerte des Zyklonabscheiders Aufgrund Meuer and
Verbessester
Messenger-Chemie-Ing.-Techn., Vol. 42, No. 5, 1970

8. Muschelknautz
The Calculation from Cyclone Separator for Gas (In German)
Chemie-Ing.-Techn., Vol. 44, No. 1 and 2, 1972

9. Muschelknautz
Centrifugal Separator Studies
Chem. Ingr. Tech., Vol. 39, No. 5 and 6, 1967

10. Muschelknautz
Design of Cyclone Separators in the Engineering Practice
Staub-Reinhalt, Vol. 30, No. 5, 1970

11. Shepherd and Lapple


Flow Pattern and Pressure Drop in Cyclone Dust Collectors
Industrial and Engineering Chemistry, Vol. 31, No. 8, 1939

12. Shepherd and Lapple


Flow Pattern and Pressure Drop in Cyclone Dust Collectors
Industrial and Engineering Chemistry, Vol. 32, No. 9, 1940

A-85
13. Stairmand
Design and Performance of Cyclone Separators
Fluid Handling, March 1952

14. Stairmand
Pressure Drop in Cyclone Separators
Engineering, October 21, 1949

15. Tengbergen
Vergleichsuntersuchungen an Zyklonen (In German)
Staub, Vol. 25, No. 11, 1965

SEPARATOPsS

The following are additional papers on steam separators that were not summarized

1. Bangma
The Development and Performance of a Steam-Water Separator for Use on
Geothermal Bores
New Sources of Energy, Proceedings of the Conference, 1961

2. Hopkinson
A New Model for the Dynamics of Steam Drums
AEEW-M1123

3. Kutepov
Centrifugal Separation of Secondary Vapors in Evaporation of Solu­
tions
Internal Chemical Engineering, Vol. 2, No. 3, July 1962

4. Mansson, Nilsson, Noberius, and Sundguist


Steam Separating Means for Boiling Water Reactors
U.S. Patent 3821078

5. Franz Meier
Steam Generator
U.S. Patent 3894517

6. Moen, Kudirka, and Gifford


Radial Vane Steam Separator
GEAR 5283

7. Ongman
.Steam Water Testing of Two Types of Combustion Engineering Incor­
porated Separators
GEAP 3501

8. Sulzer Brothers Limited


Improvements in and Relating to Water Separators for Saturated-
Steam Generators
British Patent 1189253

9. Wilson, Grenda, and Yant


Separation of Water from Steam in a Boiling Water Reactor
ASME Paper 63-WA-250

A-86
10. Polyakov
Cyclone Separators for Gas-Liquid Mixtures
Chemical and Petroleum Engineering
110.5-6, May-June 1968

RELATED AREAS

Extensive research has gone into the development of fractionating devices such as
bubbles caps and scrubber systems. Some of this technology possibly could be
applicable to the design of separators and dryers. Applicable references are:

1. Calvert, Goldshmid, Leith and Mehta


Wet Scrubber System Study Volume 1
Scrubber Handbook
EPA-R2-72-118a

2. Winkle
Distillation
Chapter 12, Fractionation Devices
MeGraw Hill, 1967

SURVEYS

Several of the references reviewed were detailed surveys of steam separation equip­
ment. These were too comprehensive to be summarized in a few paragraphs. The first
reference covers natural separation, inertia separators, spray cyclones, and
flooded cyclones. In the second reference, Gardner reviewed the performance of
knitted wire mesh and corrugated plate separators. In the third reference, Rouhani
discussed the performance of several upcomer and downcomer separators, steam dryers
and natural separation.

1. Gardner
Separation of Liquids from Gases or Vapors
HTFS Design Report No. 46, 1977

2. Ginoux
Two-Phase Flows and Heat Transfer
Chapter 13, Steam-Water Separation (Rouhani)
Hemisphere Publishing Company, 1978

3. Moore and Siererding


Two-Phase Steam Flow in Turbines and Separation
Chapter 7, External Water Separators (Gardner)
Hemisphere Pub!ishing Company, 1976
WIRE MESH DEMISTERS

Wire mesh demisters are used in the Process, Industry, and some of the early fossil
boilers, BWR's and PWR's. Because of the following two drawbacks, wire mesh is not
widely used in the Nuclear Industry today.

1. Potential for corrosion

2. Corrugated chevrons have siightly higher allowable inlet flow velo­


cities.

Following is a list of 1 iterature related to wire mesh demisters. Literature in


this 1ist was not summarized.

1. Alam and Agarwal


Design Your Gas-Liquid Separator
Chemical Age of India, June 1969

2. Bradie and Dickson


Removal of Entrained Liquid Droplets by Wire Mesh Demisters
Proc. Instr. Mech. Engrs., Vol. 184, Part 3C, 1969-1979

3. Burkholz
Drop Separation of Wire Filter (In German)
Chemie-Ing. Techn. 42, Vol. 42, No. 21, 1970

4. Burkholz and Muschelknautz


Drop Separator; Survey for the State of the Knowledge (In German)
Chemie-Ing.-Techn., Vol. 44, No. 8, 1972

5. Carpenter and Othmer


Entrainment Removal by a Wire-Mesh Separator
A.I.Ch.E Journal, Vol. 1, No. 4, December 1955

6. Dickson and Morrison


The Performance of Wire-Mesh Demisters
Hariot-Watt University, Edinburgh, England

7. Kerns
New Charts Speed Drum Sizing
Petroleum Refiner, Vol. 39, No. 7, 1960

8. Loffler and Muhr


Die Abscheidung von Feststoffteilchen Und Tropfen an
Kreiszylindern
Infolge Von Tragheitskraften (In German)
Chemie-Ing.-Techn., Vol. 44, No. 8, 1972

9. May and Clifford


The Impaction of Aerosol Particles on Cylinders, Spheres, Ribbons
and Discs
Ann. Occyp. Hyg., Vol. 10, 1967

A-88
10. Morsi
Impingement Separators - An Experimental and Theoretical Investi­
gation
Filtration and Separation, September/October 1975

11. Niemeyer
Check These Points When Designing Knockout Drums
Hydrocarbon and Processing and Petroleum Refiner, Vol. 40, No. 6,
June 1961

12. Peterson and Whitby


Fractional Efficiency Characteristies of Unit Type Collectors
ASHRAE Journal, May 1965

13. Scheiman
Horizontal Vapor-Liquid Separators
Hydrocarbon Processing and Petroleum Refiner, Vol. 43, No. 5,
May 1964

14. Scheiman
Size Vapor-Liquid Separators Quicker by Nomograph
Hydrocarbon and Processing and Petroleum Refiner, Vol. 12, No. 10,
October 1968

15. Sherwood, Shipley and Holloway


Flooding Velocities in Packed Columns
Industrial and Engineering Chemistry, Vol. 30, No. 7, 1938

16. Stairmand
Dust Collection by Impingement and Diffusion
Meeting of the Midland Branch of Institution - University of
Alabama, 1950

17. Wong and Johnstone


Collection of Aerosols by Fiber Mats
Engineering Experiment Station, University of 111inois
Technical Report, No. 11, October 31, 1953

18. York
Performance of Wire-Mesh Demisters
Chemical Engineering Progress, Vol. 50, No. 8, 1954

19. York and Poppele


Wire Mesh Eliminators
Chemical Engineering Progress, Vol. 59, No. 6, June 1963

20. Younger
How to Size Future Process Vessels
Chemical Engineering, 1955

A-89
APPENDIX B

PRESENTLY USED SEPARATOR EVALUATION


ILLUSTRATIONS

Figure Page
B-l Design 1 Separator B-3
B-2 Design 1 Separator Top View B-4
B-3 Primary Carryover Versus Circulation Ratio for B-5
Steam Flow Rate of 32000 Lbm/hr
B-4 Circulation Ratio Versus Steam FIow Curve B-5
B-5 Pressure Loss Versus Steam Flow Rate Curve B-5
B-6 Overall Carryover Versus Steam Flow Rate Curve B-5
for Circulation of Ratio 6
B-7 Primary Moisture Carryover Versus Steam FlowRate B-5
B-8 Design 2 Separator B-9
B-9 Design 2 Separator Top View B-10
B-10 Design 3 Separator B-14
B-ll Axial Vanes B-15
B-12 Straightening Vanes B-15
B-13 Total Carryover B-16
B-14 Carryover from the Region Surrounding the Separator B-16
B-15 Carryover from the Separating Barrel B-16
B-16 Effect on Inlet Quality B-16
B-17 Effect on Water Level B-16
B-18 Effect of Steam Velocity B-16
B-19 Effect of Inlet Quality on Overal1Carryover and Carryunder B-17
B-20 Effect of Water Level on OverallCarryoverand Carryunder B-17
B-21 Design 4 Separator B-20
B-22 Design 4 Separator Upper Riser and Skimmers B-21
B-23 Desi gn 4 Separator Watch Catch B-22
B-24 Desi gn 4 Separator StraighteningVanes B-23
B-25 Design 4 Separator Water Seal and Restrictor B-24
B-26 Design 4 Separator Vanes B-25
B-27 Carryunder Versus Water Level B-26
B-28 Moisture Behind Separator Versus Water Level B-26
B-29 Combined Separator-Dryer Characteristics B-27

B-i
B-30 Design 3 Separator Centrifugal Vanes B-30
B-31 Design 5 Separator Centrifugal Vanes B-31
B-32 Design 5 Separator Screen Assembly B-32
B-33 Carryover for a Family of Water Flow Rates B-33
B-34 Carryunder for a Family of Water Flow Rates B-34

B-ii
Appendix B

PRESENTLY USED SEPARATOR EVALUATION

This section presents a review of the design features of each of the five presently
used nuclear steam generator separators which were evaluated in detail. Particular
designs were evaluated, but the reader is cautioned that each manufacturer may
produce other designs. A verbal and graphic explanation of how the separators work
is presented along with sketches of each separator design. A discussion of the
performance of each separator, based upon available data, is included.

Designs 3, 4, and 5 are those recommended for possible further development for rea­
sons discussed in Appendix F.

DESIGN 1 SEPARATOR (CURTIS WRIGHT)

Figures B-l and B-2 are schematic drawings of this separator. Numoers on the sketch­
es refer to features identified in the text below. The steam and water mixture from
the tube bundle of the steam generator flows through the inlet of the separator (1)
and is diverted 90° by a top plate to the curved arm vanes (2). The steam and water
are separated after leaving the vanes. The water is slung against the downcomer
wall and flows downward to return to the tube bundle. The steam rises between the
vanes and travels to the dryers which remove most of the remaining moisture.

The function of each numbered feature on the sketches is described below:

1. Inlet to Separator - In the service application one separator is


used in a steam generator. Because the inlet is therefore an ex­
tension of the tube bundle shroud in this application, there is no
shock loss at the inlet. The tube bundle is located directly
below the inlet to the separator.

2. Curved Arm Vanes - The steam and water mixture flow through the
vane arms which give a rotating motion to the mixture causing the
water to form a vortex on the downcomer wall. The steam rises in
the annulus between the water vortex on the downcomer wall and the
riser wall. If the arms are too close to the wall, then the steam
will exit into the water vortex and results in excessive steam
entrainment. By the arms not being too far from the downcomer
wall, the water exits tangentially into the water vortex and
minimizes splashing (Reference 1).

B-l
3. Downcomer- The water level should be above the bottom of the down­
comer and riser annulus to prevent the separated water from spill­
ing into the water pool causing excessive splashing which could
lead to additional entrainment.

Warranty condition for most nuclear steam generators limits carryover from the steam
exit nozzle to no more than 0.25% by weight. Figures B-6 and B-7 show that for a
steam flow rate of 22000 Lbm/hr, system pressure of 345 psi, and circulation ratio
of 6, the overal1 carryover is 0.2%, and the carryover at the exit of the separator
is 23%. Warranty condition appears to be the upper 1imit for which this curved vane
separator operates. Both primary and secondary carryover improve substantially as
steam flow rate decreases. (References 8 and 10).

Water flow to the separator has a significant effect on the carryover performance.
Figure B-3 shows that below a circulation ratio of 7, carryover is a relative con­
stant value of 3.0%. Above a circulation ratio of 7, carryover increases rapidly.
For example, at a circulation ratio of 7.5, carryover from the separator is 6%, and
for a circulation ratio of 8.5, carryover is 25%.

Carryunder increases as water level increases. Test results for a similar separator
described in Reference 1 shows this. The reason for the increase in carryover is
that the water vortex thickness inside the separator increases as water level in­
creases and this floods the curved arm vanes.

Reference 10 indicated that for a steam flow rate of 22000 Lbm/hr and a circulation
ratio of 9, carryunder is 0.2%. Carryunder should decrease with decreasing water
flow rate and/or with increasing water level.

Figure B-4 shows how the circulation ratio for the steam generator described in
Reference 10 increases with decreasing steam flow rate. Figure B-5 shows how the
pressure drop acorss this separator increases with increasing steam flow rate.

B-2
Inner Wal1

4.04"

Steam Generator
Wall

Water Level
Downcomer Wall

Riser Wall

Figure B-l. Design 1 Separator

B-3
B-4
Circulation Ratio
Figure B-3. Primary"Carryover Versus Circulation
Ratio for Steam Flow Rate of 32000 Lbm/hr

o c; 0.4

Steam Flow - Lb/hr x 1000 12000 14000 16000 18000 20000 22000 24000
Figure B-4.. Circulation Ratio Versus Figure B-6. Overall Carryover Versus
Steam FIow Curve Steam Flow Rate Curve for Circulation
Ratio of 6.

1T000 1^000 1^000 L8000 2000 "22600" 24000

Figure B-5. Pressure Loss Versus Steam Figure B-7. Primary Moisture Carryover
Flow Rate Curve. Versus Steam Flow Rate.
Performance Curves for Design 1 Separator

B-5
DESIGN 2 SEPARATOR (WESTINGHOUSE)

In the service application, three of the Design 2 separators are located on a sup­
port plate above the tube bundle of a steam generator. Sketches of this separator
are shown in Figures B-8 and B-9. The steam and water mixture enters through the
separator inlet (1) and flows through the turbo vanes, a water vortex is formed on
the inner wall with a steam vortex in the center. The water is redirected into
the discharge channel (3) by the 1 ip of the exit orifice (5) and annular top plate.
The steam exits through the exit orifice (5). Some steam and water mixture at
the interface flows through the tangential ports (4) which siing the water against
the shell of the steam generator allowing the steam to rise upward. The steam
then flows through the dryers, which extract most of remaining moisture, before
entering the turbines. The water flows into the downcomer of the steam generator
and circulates across the tube bundle with the feedwater. (Reference 2)

The function of each numbered feature on the sketches is described below.

1. Separator Inlet - This channels the steam and water mixture from
the steam generator riser to the separator.

2. Turbo Vanes - The vanes appear to be straight. The angle of the


vanes create the centrifual force. The large size of the vanes
and therefore large flow area minimizes the shock loss.

3. Discharge Channel - The separated water flows along the inner wall
of the separator and is directed 180° into the discharge channel
by the top plate and 1 ip of the exit orifice. From the discharge
channel the water flows into the downcomer of the steam generator
and recirculates.

4. Tangential Ports - The two tangential ports separate the water


from the moisture laden steam in the transitional region where the
water enters the discharge channels and the steam flows into the
exit orifice. The water is spun into the chamber between the
steam generator shell (not shown) and the separator. The steam
flows upward and the water flows into the steam generator down­
comer (Reference 2).

5. Exit Orifice - The orifice size is chosen to optimize carryunder


and carryover performance. The orifice has to be small enough to
force the water vortex into the discharge channel; otherwise,
excessive carryover would occur. If the orifice is too small,
steam would be forced into discharge annulus (higher carryunder).
Too small an orifice would result in steam jetting out, and this
would reduce the gravity separation above the separators.

6. Hub - The vanes are mounted on the hub. The hub-separator diameter
ratio is chosen to optimize separation.

B-6
Reference 7 shows how an analytical technique (momentum flux model) can be used to
predict the carryover performance of a Design 2 type separator. Reference 8 uses
this technique and predicts the carryover performance of this separator. The sep­
arator efficiency was calculated to be 0.753 and exit quality 61.8% for the follow­
ing conditions:

Saturation Pressure - 785 psi

Steam Flow per Separator - 1,167,000 Lbm/hr

Water Flow per Separator - 2,916,670 Lbm/hr

Water Level - 43" below top of separator

Reference 8 indicates that as circulation ratio (water flow to the separators) in­
creases, separator efficiency increases siightly, and exit quality increases
sharply.

CR Efficiency Exit Quality

2.5 .743 34.0%

3.5 .753 61.8%

4.5 .763 70.2%

A similar analysis was made to see the effect :steam flow had on the exit quality
performance.

Water Flow Lbm/hr Steam Flow Lbm/hr Efficiency Exit Quality

2916600 583300 .60 34.0%

291660 1167000 .75 61.8%

291660 1750000 .75 70.2%

As discussed earlier, increasing the operating water level of the separator results
in a poorer carryover performance, and an improvement in carryunder performance.
An analysis (References 7 and 8) showed that increasing the water level does have a
detrimental effect on carryover.

CR Height Efficiency Exit Quality

3.5 30" below reference evaluation 0.753 61.8%


(low water level)

3.5 15" below reference evaluation 0.623 51.5%


(high water level)

B-7
No information is available on the carryunder performance of the Design 2 separator.
High water flow rate and water level should have an adverse effect on carryunder on
this separator. Maximizing the diameter of the exit orifice will tend to improve
carryunder performance at the expense of carryover.

The analytical technique described in References 7 and 8 probably gives the correct
trends for this separator's performance. The exit quality values predicted by the
analytical technique for this separator may be too low. Performance test data for
this separator was not available.

B-8
47.5
Annular Top Plate
Steam

Steam and Water

Inner Wall

Steam and Water Mixture

Figure B-8. Design 2 Separator

B-9
Figure B-9. Design 2 Separator Top View

B-10
DESIGN 3 SEPARATOR

Figures B-10 through B-12 describe the Design 3 separator. The steam and water
mixture enters the separator standpipe from the core of the reactor vessel. The
mixture then flows through the axial vanes (1) and is separated leaving a water
vortex on the inner wal1 and a steam vortex in the center. The primary skimmer (2),
secondary skimmer (3), and tertiary skimmer (4) skim successive layers of the water
vortex. The steam exits upward through the center of the separator. (Reference 3)

The function of each numbered feature on the sketches is described below:

1. Axial Vanes - The axial vanes are located in the conical annular
plenum between the hub and the inner barrel. The vanes appear to
be axial which would give the flow more of a upward velocity com­
ponent than other designs. The flow has less of an outward turn
and thus less shock loss. The vanes give the two phase mixture an
upward rotating motion resulting in a water vortex being formed on
the inner wal1 with the steam in the center of the separator.

2. Primary Skimmer - A large portion of the water vortex on the wall


passes through the primary discharge skimmer (Reference 3).

3. Secondary Skimmer - Steam and water passes through the secondary


skimmer.

4. Tertiary Skimmer - The remaining water and some steam is skimmed


off by this final skimmer.

5. Restriction Ring - The restriction ring is sized to optimize the


water vortex thickness on the inside barrel so that the primary
skimmer can "pick-off" the majority of the water, and the secondary
skimmer can skim the steam/water interface. (Decreasing the
clearance at the restrictor ring creates back pressure which in­
creases the water vortex thickness on the inner wal1 of the sep­
arator. ) The restriction ring causes the lowest annular flow
passage to be fi11ed with water to prevent steam entrainment and
resulting carryunder. The restriction ring is water flow depend­
ent (Reference 4).

6. Straightening Vanes - These arrest the rotary motion of the water


and reduce the velocity at the discharge.

7. Hub - The vanes are attached to the hub. It is designed to accel-


erate the steam and water mixture.

The nominal design total flow rate and inlet quality for this concept are 405,000
Lbm/hr and 14.85% respectively. For about 30" of water level, the carryover per­
formance of the separator is 2.5%, and the carryunder performance is approximately
0.25%. Increasing the total flow results in a total flow rate of 500,000 Lbm/hr
(Reference 4).

B-ll
The total carryover from the separator is generated from two sources: Water en­
trained from the free surface region surrounding the separator and carryover from
the separating barrel. The carryover from the free surface region surrounding the
separator increases with increasing steam flow rate. As the steam flow rate in­
creases , more and more steam flow exits through the secondary and tertiary dis­
charge passages, resulting in a sufficient high velocity between the separators
to entrain and carry water with it. Figure B-14 shows the carryover dependency
from this source in terms of steam flow rate, total flow rate, and water level.
Carryover from the separating barrel increases exponentially when the inlet quality
drops below 11% (Figure B-15). As the inlet quality is decreased the vortex thick­
ness increases, and at some low quality there is a step-type transition to a very
thick vortex. Carryover in this case results because the water vortex becomes too
thick to be picked-off by the skimmer rings. The total carryover from both sources
correlated as a function of inlet quality and water level. As inlet quality dec­
reases below 10% or increases above 16%, total carryover increases sharply (Figure
B-13).

Reference 4 indicates that carryunder comes from these sources:

1. Steam passing through the lower annular flow discharge passage


because of incomplete separation.

2. Steam passing through the lower annular flow passage as result of


vortex thickness being smaller than skimmer gap.

3. Steam entrained from free surface around separator by jet action


of one or more of the discharge passages.

Figures B-16, B-17, and B-18 illustrates how quality, water level, and water flow
rate, respectively, affect the carryunder performance. Figure B-16 shows the effect
inlet quality and restrictor ring clearance has on carryunder. For high qualities,
the vortex thickness in the barrel decreases and carryunder increases due to the
increasing amount of steam that is skimmed off by the skimmer rings. For low qual­
ities , the vortex thickness increases resulting in more water passing through the
secondary and tertiary discharge passages. The increases in carryunder for low
qualities results from the jet action of water flow resulting in splashing from the
secondary and tertiary discharge passages.

The restriction area in the primary discharge channel is sized to optimize carryunder
performance at design flow conditions. The restriction clearance is probably water
flow sensitive and affects the vortex thickness. A tight (small) clearance might
tend to decrease carryunder while increasing carryover. On the other hand, a loose

B-12
(large) clearance would have a detrimental affect on carryunder and improve carry­
over.

Figure B-17 shows how carryunder increases with decreasing water level. Reference
4 shows that the source of this carryunder is the steam passing through the primary
discharge passage becuase the water vortex thickness is less than the skimmer ring
gap.

Figure B-18 is a plot of carryunder versus the superficial liquid velocity in the
region outside the separator. High water flow rates correspond to high superficial
liquid velocities. At the high liquid velocity, less steam, that is carried under
from the separator, is allowed to become de-entrained due to the greater drag force
on the steam bubbles.

The overall carryunder and carryover performances are illustrated on Figures B-19
and B-20 as a function of inlet quality and water level. The overall carryover is
the carryover from the dryers.

B-13
Steam

Steam

Steam |

Hater

Steam

Water
Steam

Water

Steam

Water

Steam and Water Mixture

Figure B-10. Design 3 Separator.

B-14
Figure B-ll. Axial Vanes

Figure B-12. Straightening Vanes

B-15
" L=10

Af=14.9 in
0 0.2

8 10 12 14 16 18

O 10 Inlet Quality x-%


Figure B-16. Effect of Inlet
Quality.

Wt=640,000 Lb/hr
x=14%
Inlet Quality x-% sAf=14.9 in^
Figure B-13. Total Carry­
over. o 0.2
455000<Wt<640/000 Lb/hr
10 20 30 40 50

Water Level L-inches


Figure B-17. Effect of Water
Level.

00 120 140 460,000<Ws<640,000 Lb/hr

Steam Flow Rate Ws ~ Lb/hr


Figure B-14. Carryover from
the Region Surrounding the
Separator
45 ,000<Wt<640,000 Lb/hr

8 .12 1.6
Inlet Quality x-%
Liquid Velocity V$ - ft/sec
Figure B-15. Carryover from
Figure B-18. Effect of Steam
the Separating Barrel
Velocity

B-16
Wt = 405.000 !b/h

X - 14.3 %

L * 25 in.

-lO 0 10 20 2C -iO 3*?

WATER LEvcL. L-m.


iNLHT QUALITY. X -

Figure B-19. Effect of Inlet Figure B-20. Effect of Water


Quality on Overall Carryover Level on Overal1 Carryover and
and Carryunder. Carryunder.

Overal1 Carryover and Carryunder for Design 3 Separator

B-17
DESIGN 4 SEPARATOR

In the service application of this design a steam and water mixture rises from the
steam generator tube bundle and enters several steam separators mounted on a deck
plate. Figures B-21 through B-26 describe the mechanical features of this
separator. Referring to Figure B-21, the steam/water mixture enters the lower
riser (1) and is then diverted by the top plate through the swirl vanes (2). A
water vortex is formed on the inner surface of the outer wall with a steam vortex
near the outer surface of the inner wall. As controlled by the clearance at the
restrictor (8), portions of the water vortex flow both upward and downward. The
portion that flows upward is "picked-off" by the lower skimmer (5) and upper
skimmer (6) and flows down the outer wal1 of the separator. The portion of the
water that flows downward goes through the water seal (9) and passes around the
restrictor (8) to the separator support plate. Curvilinear channels are
provided between the swirl vanes to allow part of the steam to flow into the upper
riser. The steam flow is then diverted horizontally by the water catch (4) to the
space above the separator. The other portion of the steam flow rises in the
annulus between the inner and outer walls and flows by the drain pipes from the
water catch. Moisture collected by the water catch flows down the drain pipes to
the outer wall of the separator.

The function of numbered features on the sketches are described below:

1. Lower Riser - The steam and water is channeled to the swirl vanes
via the lower riser (See Figure B-25).

2. Swirl Vanes (See Figure B-26) - These vanes are formed by alternated
one piece curve member and curvillinear channels. Two phase flow
flow enters these vanes from the riser. By centrifugal action of
the vanes the steam and water are separated. The water forms a
swirling film on the inner side of the outer wall. Part of the
steam exits through the curvilinear channels and the other part
rises in the annulus.

3. Upper Riser (See Figure B-22) - After the steam and water is
separated by the swirl vanes, part of the steam flows through the
curvilinear channels and exits through the upper riser.

4. Mater Catch (See Figure B-23) - The water catch distributes the
steam flow throughout the area above the separator. This reduces
jetting and allows for more gravity separation. Water droplets
adhere to the bottom of the water catch. Drainage pipes are
provided to channel the water to the outer wal1 of the separator
(Reference 5).

5. Lower Skimmer (See Figure B-22) - Part of the water film traveling
in the upward directi on is skimmed off. The thickness of the water
film is a function of the restrictor clearance and water flow rate.

B-18
6. Upper Skinner (See Figure B-22) - The remaining part of the water
film is skimmed off by the upper skimmer.
7. Straightening Vanes (See Figure B-24) - The straightening vanes
remove the swirl from the water flow exiting the skimmer and
reduce the velocity.
8. Restrictor (See Figure B-25) - Varying the restriction clearance
throttles water upward and downward along the inner surface of the
outer wal1.
9. Water Seal (See Figure B-25) - A water seal is maintained at the
exit of the separator downcomer.

Carryover is a strong function of water level. For a pressure of 1000 psi, steam
flow rate of 7500 Lbm/hr, and water flow rate of 309000 Lbm/hr, Figure 28 shows
carryover to be about 1% up to a water level of 2.65 ft. At the normal water level
2.95 ft. carryover is 1.5%, and above 2.95 ft. carryover increases rapidly.

Figure B-29 shows that the carryover performance of this separator is rather in­
sensitive to steam flow rate. At a steam flow rate of about 8000 Lbm/hr of steam,
there is a siight increase from 1.5% to 3% carryover. At all other steam flow
rates, the carryover is fairly constant at 1.5%.

No information was given in Reference 9 showing the effect of water flow on the
carryover performance on Design 4 separator. For a similar separator in Reference
5 (designated as the compact separator), there is a dependency on water flow rate.
For the compact separator at a water level of 3.28 ft. and steam flow rate of 83000
Lbm/hr, carryover increases from 2% for a water flow rate of 80000 Lbm/hr to about
10% at a water flow rate of 160000 Lbm/hr. Combustion Engineering, Inc. and others
have tested separators with restrictions in the annulus, and found them to be very
sensitive to water flow rate. For this reason, the Design 4 separator carryover
performance should be dependent on water flow rate. Higher wave flow rates increase
carryover.

For normal operating conditions, carryunder is approximately 0.2%. At 0.65 ft. of


water level on the separator, carryunder varies from 2% at 1110 psi to 0.5% at 740
psi. For all three of these system pressures, shown on Figure B-27, carryunder
gradually deveased to negligible amounts at the higher water levels.

Reference 9 does not indicate the effect water flow rate or steam flow rate has
on carryunder performance. If water flow rate and steam flow rate have the same
trends as similar designs with restrictions in the annulus, then carryunder will
gradually increase as steam flow rate increases, and decreases as water flow rate
increases.

B-19
Steam
Steam

Water

Water
Steam

Steam
Top Plate
Water Level

Steam & Water


"/Water
“ Mi v +■ 11 v'Ci

Inner Wall

Outer Wall

Water
Steam & Water
Mixture

6.7

19.3"

Figure B-21. design 4 Separator

B-20
Water

Water

Water

Steam Steam

Hater

Figure B-22. Design 4 Separator Upper Riser and Skimmers

B-21
Water
B-22

Steam

Waten

Water

Figure B-23. Design 4 Separator Water Catch


Top View

End View of
Two Vanes

Figure B-24. Design 4 Separator Straightening Vanes

B-23
Steam Water Water

Figure B-25. Design 4 Separator Water Seal and Restrictor


B-25

Figure B-26 Design 4 Separator Vanes


2.8

2.4

------------ps 740 psia


2.0
_____ Ps 925 psia
%
c ___ Ps 1110 psia
1.6
s_
<3 ____ Ps=740 psia
B-26

^ 1.2
-------- Ps=925 psia
W$ =93000 Lbm/hr
W =309000 Lbin/hr 5 ___Ps=1110 psia
0.8. \ w
\
\
W = 75000 Lbm/hr
\
0.4 W = 309000 Lbm/hr
110% load CR=4.35
■ w_________

0.66 1.97 3.28 h(FT)


1.31 2.62 3.94 h(FT)

Figure B-27. Carryunder Versus Water Level Figure B-28. Moisture Behind Separator Versus
Water Level

Performance Curves for Design 4 Separator


CARRYOVER

cH5

10
Normal Water Level: 2.95 ft

Separator Characteristics

_L_™--------------------------------- i™

0 20 30 ,40 50 60
Steam Flow Rate x 10;^ Ltp/hr Load
_
LOAD 100 110

Figure B-29. Combined Separator-Dryer Characteristics

B-27
DESIGN 5 SEPARATOR

In its service application, numerous Design 5 separators are mounted on standpipes


welded to a deck plate above the steam generator tube bundle. Figures B-30 to B-32
describe the features of this separator. Steam and water mixture enters the inlet
nozzle (1) and is redirected by the top plate to the vanes (12). The vanes impart
a centrifugal force to the mixture. Water flows in a vortical motion up the inner
wall of the separator, and the steam rises in a vortical motion in the center of the
separator. The water exits through the perforations (4 and 5) in the wall, and the
steam flows through the exit nozzle (6) into the screen assembly (7), then out of
the separator. Some of the moisture that was entrained in the steam is extracted
by the screen assembly and flows down the outer wall of the separator. (Reference 6)

Numbered features on the sketches are described below:

1. Inlet Nozzle - The inlet nozzle is adapted for mounting upon the
upper end of a standpipe. Also, it channels steam and water flow
to the vanes.

2. Centrifugal Vanes - The centrifugal vanes impart a high rotary


motion to the two phase mixture whereby the 1iquid phase of the
mixture is thrown radially outwardly toward the inner surface of
the separator wall and is thereby separated. The vapor flows up­
wardly through the central portion of the separator.

3. Barrel - A water vortex is formed inside the separator. The film


thickness of water becomes thinner with increasing height of the
barrel as water exits the separator through perforations in the
barrel. At higher water flow rates and/or water level external to
the separator, the height and thickness of the film increases.

4. Lower Perforations - The diameter and spacing of the perforations


are selected to present an open wall area that is sufficient to
pass a major portion of the separated liquid transversing the
surface of the separator wall while, at the same time, being not
so great as to disrupt the vortical character of the following
liquid or permit leakage of significant amounts of vapor through
the separator wall.

5. Upper Performations - The diameter and spacing of the perforations


in the upper portion are selected to present an open wall area that
is sufficient to pass essentially all of the remaining separated
1iquid while maintaining the vortical character of the mixture
within the barrel.

6. Discharge Nozzle (Skimmer) - The skimmer serves as a barrier to


impede the upward flow of any 1iquid film that may exist in this
region of the chamber. A discharge opening that is too smal1
results in carryunder, and an opening that is too large results
in carryover.

B-28
7. Screen Assembly - The screen assembly distributes the flow across
the rise area. Jetting from the outlet nozzle is prevented. Jet­
ting of the steam directly into the downstreat dryer would reduce
its drying capacity.

The screen assembly also extracts a small amount of moisture. The water that is
collected by the screen assembly drains down the outer surface of the barrel. The
velocity of the steam through the screens is too high for maximum drying.

For the Design 5 separator, carryover varies from a negligible amount at a stem
flow rate of 35000 Lbm/hr to about .2 to .5% at 60000 Lbm/hr. For nominal condi­
tions, a change in water flow rate of 40000 Lbm/hr to 210000 Lbm/hr will increase
carryover from 0.02% to 0.15%. Up to a certain point, an increase in water level
results in a slight increase in carryover, and past that point the increase is
rapid. For high water and/or steam flow rates this increase in carryover occurs
at lower water levels. Figure B-33 is a typical plot illustrating how carryover
varies with water level and water flow rate for a nominal steam flow rate.

Water flow rate and water level have a pronounced effect on carryunder, and steam
flow rate only has a slight effect on carryunder. As water level increases from 4"
below the vanes to the vanes, carryunder decreases from 4% to below 1%. For higher
water levels and as water flow rate decreases in equal increments from 210000 Lbm/hr
to about 120000 Lbm/hr, carryunder decreases in larger and larger increments. Below
120000 Lbm/hr, carryunder sharply decreases to negligible amounts. Figure B-34 is
a typical plot illustrating how carryunder varies with water level and water flow
rate for a nominal steam flow rate.

B-29
11.25
4 Screen Assembly

Steam

1/8" 0 Perforations
1/4" triangular
pitch

Water,

5/16" 0 Perforations
3/4" x 1" square
spacing

GO_______

Top Plate

Figure B-30. Design 5 Separator Centrifugal Vanes

B-30
SECTION A-A

Figure B-31. Design 5 Seoarator Centrifugal Vanes

B-31
7\7\

Figure B-32. Design 5 Separator Screen Assembly

B-32
Water Flow Rates
1 - Nominal Value , r
1.6

'J I-------- > 1--------1------- 1—J_ U 4 1 i- i —I -1 !. I I J .. . i. .1 I J 1 1.. . I...

0 1 2

Water Level (1-nominal value)

Figure B-33. Carryover for a Family of Water Flow Rates.

B-33
Water Flow Rate
(1-Nominal Value)

Water Level (1-nominal value)

Figure B-34. Carryunder for a Family of Water Flow Rates.

B-34
REFERENCES

1. R. C. Marshall. "Centrifugal Liquid-Vapor Separation." U.S, Patent 3507099.

2. P. F. Sokolowski. "Vertical Steam Drum." U.S. Patent 3751886. 1973.

3. R. H. Moen and S. Wolf. "Gas-Liquid Vortex Separator." U.S. Patent 3902876.


1975.

4. S. Wolf and R. H. Moen. "Advances in Steam-Water Separators for Boiling


Water Reactors." ASME 73-WA/PWR-4. 1973.

5. H. Brandes. "Experimental Investigation of Centrifugal Separators-1."


Energie and Technik. Vol. 23, 1971, pp. 283-290.

6. Patterson, Micheller, and Carney. "Centrifugal Separator Having Perforated


Can Wall." U.S. Patent 3614863. 1968

7. R. R. Bennett and N. N. Kondic. "Momentum FIux Model for Liquid-Vapor


Separation." In Proceedings of 1977 Heat Transfer Conference. 1977.

8. D. A. Steininger. "EPRI Correspondence to W. R. Carson." 1978.

9. R. Bouecke and G. Schucktanz. "Two Phase Flow Separation in U-Tube Steam


Generators." In Proceedings of 17th National Heat Transfer Conference,
1977.

E. B1urn, A. Sievers, and D. Wagner. "Steam Generator Design and Development


Capability." Curtis Wright Corporation Report. 1977.

B-35
Appendix C
DRYER DESIGN FEATURES AND PERFORMANCE
ILLUSTRATIONS

Figure Page
C-l Vertical Baffle Dryer C-4
C-2 Corrugated Plate Chevrons C-5
C-3 Delta P C-6
C-4 Euroform C-6
C-5 Peerless C-7
C-6 Westinghouse C-7
C-7 King Tool C-8
C-8 Peco C-8
C-9 Burgess Manning C-9
C-10 Steinmuller C-9
C-ll Hooked-Vanes (Flow Over Scoops) C-10

C-i
Appendix C

DRYER DESIGN FEATURES AND PERFORMANCE

This project was directed toward evaluating the performance of primary steam
separators and their affects on steam generator operation. During this project,
the secondary separators (dryers) were also evaluated since they have an affect
on overall steam generator performance.

An evaluation of three general types of dryers is presented herein. Each of the


three dryers have advantages and disadvantages. In general, wire mesh demisters
and corrugated baffle dryers have lower moisture removal capability as compared
to hooked-vane dryers. The corrugated baffle is more compact (requires less
volume) than the wire mesh demister and hooked-vane dryer. The wire mesh de­
mister is usually easy to assemble and relatively inexpensive. The dryer system
should be determined based on the separator efficiency for a given steam generator
application.

WIRE MESH DEMISTER

A wire mesh demister consists of several layers of screen wires designed to


separate moisture from steam. Separation occurs when the liquid droplets adhere
to the mesh and fall off as the force of gravity overcomes the combined forces
of the rising gas and surface tension which tends to hold the droplets on the
wire. At a threshold (critical) velocity, the bottom portion of the wire is
flooded with moisture, and massive carryover occurs because the wire mesh is not
able to drain (Reference 1). Below this threshold velocity of about 5 ft/sec at
an operating pressure of 1000 psig, moisture removal can approach 100% and
higher inlet moisture is permissible. The performance of the wire mesh improves
slightly if the wire mesh is inclined or if it is a finer mesh. The optimum
performance appears to occur at an angle of 45° (Reference 2).

Because of the low threshold velocity for wire mesh and the possibility of
corrosion and erosion, wire mesh is not recommended. Reference 2 reported that
the wire mesh demister in the San Onofre Moisture Separator Reheater in the

C-l
turbine cycle wore through locally and severed after operating only three months.
The breaks were analyzed and their cause attributed to abrasion.

CORRUGATED BAFFLE CHEVRON DRYERS

Corrugated baffle dryers consists of stacked baffle plates through which the
fluid flows with the moisture collecting on the chevrons (corrugated plates) and
draining due to gravity. Reference 3 recommends vertical or inclined baffle
dryers because they operate at higher inlet velocities and provide the same (or
somewhat lower) residual moisture than horizontal baffle dryers, raschig ring
packings, and standard meshes. In an inclined or vertical baffle dryer the
fluid flows horizontally between the chevrons. Moisture collects on the chevrons
and flows in rivulets vertically down the troughs of the corrugations. The de­
terioration of performance in the baffle dryer is associated with flooding or
choking in the dryer channels.

Several investigators recommended the vertical chevrons over the wire mesh or
horizontal baffle dryers. A moisture separator reheater is mentioned in Reference
3 where wire mesh was replaced by vertical chevrons, and the steam flow rate was
allowed to increased by 45%. The vertical or slightly inclined corrugated dryer
is documented in Reference 4 as having approximately 80% better carryover per­
formance than a horizontal baffle dryer. Reference 3 indicated that the vertical
baffle dryer had 2 to 4 times greater critical velocity than the horizontal
baffle dryer.

One vertical baffle dryer (Figures C-lb and C-2) reviewed was 6.5" high, inclined
at a 30° angle. By incl ining the dryer as shown in Figure C-l, the inlet flow
area is increased, thus allowing for slightly higher steam flow rates. For an
angle up to 30°, there is no detrimental affect on the drainage or critical
velocity of the dryer (Reference 5). The short dryer has a drainage advantage
over a taller dryer. Moisture accumulates and fills the flow passages in the
lower portion of the taller dryer resulting in more moisture carryover for some
conditions. The moisture from the short dryer (Figure C-lb) drains to a bottom
plate and either flows to drainage pipes at the exit of the dryer or fal1s into
the water pool around the centrifugal separators. By not incorporating a
drainage tray below the dryer, the potential problem of part of the steam by­
passing the chevrons through the drainage try (Figure C-la) is eliminated.

C-2
HOOKED-VANE BAFFLE DRYER

Hooked-vane baffle dryers are a variation of the corrugated baffle dryer. The
hooked-vane baffle dryer consists of stacked baffle plates in which the steam
flows between the baffles in the horizontal direction. The surface of each
plate is comprised of a section which is duplicated a number of times to form
a serpentine, or zig-zag pattern. At the top and/or bottom of each section
hooks (scoops) are provided for drainage of the liquid that collects on the
plates. In Chapter 7 of Reference 2, the operation of a dryer similar to the
ones in Figure C-5 is explained. The water does not actually drain down the
scoops. A vortex is set up within the scoops, which displaces the water and
causes it to drain as a ribbon in front of the scoops. Reentrainment, when it
occurs, is from the ribbon.

The advantage of the hooked-vane dryers shown in Figures C-3 through C-10 is
that they extract large amounts of moisture. Various references report the
upper limit to be from 10 to 30% moisture. The closed hooked-scoops similar to
the ones in Figures C-8 and C-9, should perform better than open hooks. In
these dryers, there should be less reentrainment than with the open hooks where
the water drains in front of the scoops.

According to the description of the operation of the open hooked-vane dryer


(Reference 2), the shape of the hook should not be a critical design feature.
The addition of a scoop facing the flow direction, or even allowing the flow to
transverse over the scoops similar to Figure C-ll, should serve the same purpose
in allowing the water to drain.
Drainage Tray

By Pass Flov/

b.

Figure C-l. Vertical Baffle Dryer

C-4
fig u re
DELTA-P

Ij^o ~ i ru i r ~

Figure C-3. Delta P

Figure C-4. Euroform


Figure C-6. Westinghouse
Figure C-7. King Tool

Figure C-8. Peco

C-8
Figure C-9. Burgess Manning

o
!
<JD

Figure C-10. Steinmuller


C-10

Figure C-ll. Hooked-Vanes (Flow Over Scoops)


REFERENCES

1. 0. H. York and E. W. Poppele. "Wire Mesh Eliminators". Chemical


Engineering Progress. Vol. 59, No. 6, 1963, pp. 45-60.

2. M. J. Moore and C. H. Sieverding. Two-Phase Flow in Turbines and


Separators. New York: McGraw Hill, 1976, pp. 317-369.

3. Y. L. Sorokin, L. N. Demidova, and Kuz'min. "Principles of Drop


Separation From Vapor of Gas Streams". Chemical and Petroleum
Engineering. Vol. 8, No. 4, 1968, pp. 664-668.

4. N. A. Mozharov. "Maximum Permissible Rate of Steam Flow Through a


Separator". Teploenergetika. Vol. 8, No. 4, 1961, pp. 60-63.

5. M. D. Panasenka and K. V. Koslov. "Investigation of Separating


Equipment for High Capacity Drum Boilers". Teploenergetika.
No. 8, pp. 69-72.

C-ll
Appendix D

ALTERNATIVES TO TWO STAGE SEPARATION


ILLUSTRATIONS

Figure Page
D-l Stacked Arrangement D-5
D-2 Conventional Arrangement D-6

D-3 Baffling Scheme D-7

D-i
c/
Appendix D

ALTERNATIVES TO TWO STAGE SEPARATION

Typically, separation of steam and water in a nuclear steam generator is accom­


plished in two stages. Separators are used as the first stage followed by
dryers as the second stage. In this evaluation, a third stage of separation
devices was investigated in an effort to reduce the overall separation pressure
drop by providing an alternate path for a portion of the two phase fluid. Three
different alternatives of extra stage separation devices were evaluated to deter­
mine if the pressure drop in the circulating loop of a steam generator could be
decreased without adversely affecting carryover or carryunder performance. The
three alternatives evaluated were:

1. An extra stage of separators in series,

2. An extra stage of separators in parallel, and

3. Pre-separation using baffles prior to the separators.

The second alternative, an extra stage of separators in parallel, presents the


most promise for reducing pressure drop and moisture carryover and is the only
alternative recommended for further development. A discussion of the three
alternatives is presented below.

EXTRA STAGE OF SEPARATORS IN SERIES

This alternative was eliminated because the second stage of centrifugal separators
did not perform very efficiently at the inlet moisture levels of 10% or less pro­
vided at the exit of the first stage of separators.

EXTRA STAGE OF SEPARATORS IN PARALLEL

The purpose of stacking the separators in parallel is to reduce the flow per
separator. This would produce a lower pressure drop, and for several separator
designs this would have the potential of improving carryover performance.
Separator Designs 1 and 5, described in Appendix B, exhibited a decrease in
carryover with decreased steam flow rate.

D-l
One way to arrange the separators in parallel is to place them on a square pitch
with a lower and upper level as shown in Figure D-l. If separator Design 5 were
used, an outer jacket similar to the one for separator 1 would be required to
prevent re-entrainment of the separated water. Figure D-l shows a separator
similar to Design 1 or Design 5 with an outer jacket. In order to determine the
relative increase in the number of separators possible with the stacked parallel
arrangement versus the single level conventional arrangement, the separator O.D.
for convenience was chosen to be 12" with an outlet diameter of 7" and pitch of
14". The maximum height for these separators should be kept in the 30" and 40"
range so that they wi11 fit into future steam generators without excessive in­
creases in steam drum volume. For a 3' x 4' area enclosed by the broken 1ines,
one third more separators can be placed in the stacked arrangement (Figure D-l)
than the conventional arrangement (Figure D-2). To verify adequate performance,
the proposed separator in the stacked arrangement would have to be tested. One
potential problem is that water draining from the upper stack of separators
might spill into the exiting steam resulting in carryover. A second potential
problem is that the lack of a water level on the upper stack of separators may
result in excessive carryunder.

PRE-SEPARATION USING BAFFLES AS AN EXTRA STAGE BEFORE THE SEPARATORS

The purpose of baffling (Figure D-3) in the riser of a steam generator would be
to perform gross separation at low pressure drop and channel the separated water
to the downcomer. The steam and the remaining water would go through the cen­
trifugal separators, with the pressure drop across the separator much less than
for the original flow. The advantage of this scheme is that the pressure loss
in the circulating loop of the steam generator would be less.

Figure D-3 shows two side views of the baff1ing device. Steam and water enter
the riser from the tube bundle (not shown), and flow through the baffles. The
V baffle redirects the flow, resulting in separation by impingement and by water
dropout. The water falls into the troughs of the baffles, drains along the
troughs, and spills on the inside surface of the shroud. From there, the water
slides down the shroud through slots into an annulus, composed of two shrouds,
to the downcomer.

A baffling scheme of this type may be most advantageous at partial loads. At


partial loads, the steam flow is less, but most flow codes predict the water

D-2
flow rate to be relatively constant. Lower steam flows may not be able to
carry the relatively large amount of water flow. Some of the water at partial
loads may be carried a short distance before falling back into the bundle, re­
sulting in a lower effective recirculation ratio than predicted by the codes.
If this water could be caught and channeled to the downcomer such as with a
baffling or similar scheme, then the high predicted circulation ratios might be
maintained.

The baffling scheme described is not recommended because such extensive and un­
tried modifications to steam generators do not seem to be justified by the
benefits to be obtained. The separator pressure drop is only about 20% of the
pressure loss in the circulating loop of the steam generator, and the amount of
pressure drop reduction does not significantly increase circulation ratio
(discussed in Appendix J). The baffling would hamper accessibility to the area
above the tube bundle during inspection.

An evaluation was made to determine if a third stage of separation could improve


carryover without adversely affecting carryunder. Stacking separators in paral-
lei has the potential of reducing carryover and is discussed in an earlier
portion of this appendix. Two other extra stage separation alternatives for
reduction of carryover are:

1. An extra stage of dryers in parallel and

2. An extra stage of dryers in series. A discussion of these two


alternatives follows:

EXTRA STAGE OF DRYERS IN SERIES

This alternative was evaluated in conjunction with dryers in parallel, but was
excluded from serious consideration as it was determined that no real benefit
could be derived from this arrangement. If excessive carryover from the origi-
nal stage of dryers were occurring due to steam velocity exceeding the critical
limit, the high velocity and resulting carryover would also exist in the extra
stage of dryers in series.

D-3
EXTRA STAGE OF DRYERS IN PARALLEL

If the original stage of dryers were emitting excessive carryover due to a


shortage of surface area, this arrangement would increase the dryer surface
area. This increase in area would have the potential of decreasing the carry­
over to negligible amounts if the velocity of the steam through the parallel
dryers was below the critical limit. In order to adopt this arrangement,
uniform flow distribution to the dryers and adequate drainage from the dryers
would have to be provided for in the design. Since most of the steam genera­
tors studied did not have a dryer carryover problem, this alternative is not
recommended for further development.

D-4
D-5
“ I

D-6
SI DC View
(UHJKHIO AC«Or*S
I-AHIV.S)

mfam viAirn n on

Figure D-3. Baffling Scheme.

D-7
Appendix E

DESIGN FEATURES
ILLUSTRATIONS

Figure

E-l Hypothetical Separator E-2


E-2 One Stage of Vanes E-9
E-3 Helical Vanes E-9
E-4 Rouhani's Axial Vanes E-10
E-5 Dement1 ev Axial Vanes E-ll
E-6 Cochran's Axial Vanes E-ll
E-7 Mesh Cone and Cylinder E-15
E-8 Screen Assembly E-16
E-9 Corrugated Chevrons E-17
E-10 Deflector Plate E-18
E-ll Turbo Separator with Tangential Ports E-19
E-12 Second Stage Skimmer E-20
E-13 Pre-dryer with Secondary Vanes E-21
E-14 Guide vanes E-22
E-15 Baffle Arrangement E-25
E-16 Hollow Vanes E-26
E-17 Bubble Rakes E-27
E-18 Axially Disposed Vanes E-28
E-19 Flow Channels and Vanes E-29
E-20 Separators with Restrictors and Pick-Off Rings E-32
E-21 Water Seal Separator E-33
E-22 Separator with Reverse Scoops E-34

E-i
Section 1

INTRODUCTION

Most separators are composed of one or more of the following design features:
A hypothetical separator composed of these design features is shown on
Figure E-l.

1. Vanes - They perform the initial separation of the gas and liquid.

2- Pre-dryers - In some cases excess moisture is contained in the


exiting steam. This feature removes part of this excess moisture.

3. Secondary separation designs - This feature is used to remove


part of the steam entrained in the exiting water.

4. Auxiliary parts to prevent re-entainment - After the steam and


water are initially separated by the vanes, the water and steam
has to be channeled from the separator without causing re­
entrainment.

In the following sections, the performance characteristics of variations of


these features are presented. No attempt is made to identify the optimum design
of each feature as it is the performance of the combination of features which
is important to separator design. However, the following discussion should
prove useful to the separator designer in selecting and optimizing the various
separation enhancement features.

E-l
4
Secondary Separator

Auxilary Parts To ^
Prevent Re-entrainment *

Secondary Separation
Device

Vanes

Water Steam

Figure E-l. Hypothetical Separator.

E-2
Section 2

VANES

Vanes are used in a separator to impart the centrifugal force which results in
the separation of the vapor and the liquid. This Appendix presents a summary
of some of the significant information for vanes used in nuclear steam genera­
tors. Several axial vane designs are only discussed briefly with little
technical explanation, but are included here because the writer is aware that
these designs are presently being used with success. (In these cases, the per­
formance information was not available in the reference.) Axial vanes are
usually attached to a hub which is contained in a cylinder. The two phase
mixture flows through channels composed of the hub wall, the cylinder wall, and
the surfaces of two vanes. The summaries of the references on vanes follows:

In Reference 1, four series of vane developmental tests were discussed. The


criteria used to evaluate these vanes were pressure drop and separation of the
two fluid phases. The following sub-paragraphs are a partial summary of
Reference 1 tests.

ONE STAGE OF VANES (FIGURE E-2)

Reference 1 reported that four vanes are the optimal design number. Increasing
the number of vanes resulted in an increase in pressure drop, and decreasing
the number produced a loss of water vortex (incomplete separation).

The vanes tested were 6" high and flat in shape. Of the three blade angles
tested, the 30° and 45° vanes performed adequately with respect to low pressure
drop and separation of the gas and the liquid. The 60° vanes did not produce a
large enough center vapor vortex to maintain flow through it, and therefore
did not produce the required separation.

E-3
In summary, these tests indicated that for best separation, the number of vanes
should not be less than 4. If more than four vanes were used, there would be ^n
increase in pressure drop. The optimum vane angle appears to be between 30° and
45°. It is worthy of note that the height of these vanes was 6".

TWO STAGES OF VANES IN SERIES (FLAT VANES)

Two stages of vanes were stacked in series. The best of these models had four
vanes with a first stage vane angle of 60° and a second stage vane angle of
45°.

Two stages of vanes were used in an effort to improve the design. There is a
more gradual change in flow direction to the final 45° exit angle by having the
first stage of vanes sloped at 60°. By having a gradual transition in the vanes,
there is less agitation in separation and fewer of the larger bubbles burst into
smaller bubbles resulting in more complete separation.

TWO STAGES OF VANES IN SERIES (CURVED VANES)

Two separate types of vanes were tested. The first type was composed of two
stages of vanes in series. The blades in each stage were not flat, but were
bent in increments of 10° to produce the directional change. The best perfor­
mance in this arrangement came from first stage vanes with an angle of 90°
(vertical) and second stage vanes with an angle of 50° (incremental bending).
For a water flow rate of up to 2000 gpm, carryunder averages less than 2% by
volume with this design. The second type that was tested had helical vanes
(Figure E-3) with an outside diameter of 8". The best helical design had a
twist of 135° and exit angle of 50°. For a water flow rate of 2000 gpm, the
pressure drop through this vane design was 4.5 feet of water.

In both of these tests the gradual change in vane angle served to improve vane
performance (pressure drop, separation).

THREE STAGES OF VANES IN SERIES

Three stages of vanes in series were also tested. Each stage was contained in
a 10" diameter cylinder. The vanes in each stage had 10° incremental changes
to achieve the change in direction. The arrangement with the following vane
angle for each stage had the best overal1 performance:
1. First stage - 90°

2. Second stage - 40°

3. Third stage - 40°

An arrangement of these three stages of vanes had the best overall performance.
The progression of vane tests in arriving with this design is interesting.
Their first tests were with flat vanes, and they determined that the 30° and
45° angle was optimum. Next, two stages of flat vanes were tested. Sub­
sequently, two stages of flat vanes with 10° incremental changes in the
curvature were tested. The best design in these tests had first stage vanes
which were 90° (vertical) and second stage vanes with an angle of 50°. Finally,
three stages of vanes were tested and found to have the best performance. This
design had first stage vanes of 90° and second and third stage vanes at 40° with
10° incremental changes to achieve the 40°.

In Reference 2, qualitative tests of axial vanes were described. Only photo­


graphs and visual observation provided data on the separation of air/water.
The axial vane (cyclone) assembly consisted of four curved sheet-metal vanes
contained in a 3.5" transparent acrylic plastic tube.

Three axial cyclones designated as I, II and III were tested. All three cy­
clones were tested with a 1 inch hub, and cyclone I was additionally tested
with no hub. The vanes were vertical at the inlet with a definite slope at the
outlet. The angle of the slope from the horizontal at the outlet is designated
as 0int at the hub, 0exT at the separator wall, and the shock loss factor as t .
The values of which are shown in Table 1.

Table 1. Axial Cyclones

Axial Cyclone I II III

0INT 30° 50° 50°

0EXT 10° 30° 45°

4 10.2 2.4 0.8

E-5
The axial vanes produced separation of the major part of the water in a nearly
air free condition along the wall. Air mixed with water formed a vertical cone
in the center of the pipe. There was not a clear separating surface between
the water against the wall and the mostly air cone.

The air/water cone diameter in the center of the separator was found to be de­
pendent on both inlet void fraction and superficial inlet water velocity. For
all three cyclones, it was observed that the air/water cone diameter increased
in size with the increase in inlet void fraction. Increasing superficial inlet
water velocity affected the air/water cone diameter of each of the three tested
cyclones differently.

Cyclone 1 Increasing superficial inlet water velocity increases the


air/water cone diameter.

Cyclone 2 At superficial inlet water velocities greater than 5.9


ft/sec, the air/water cone diameter increases but at lower
velocities there is essentially no effect on the air/water
cone diameter.

Cyclone III • Increasing superficial inlet water velocity up to 4.6 ft/


sec, increases the air/water cone diameter. At greater
velocities the air/water cone diameter continues increasing.

It was concluded from these tests that there is a optimum cyclone deflection
for each set of flow conditions. Furthermore, the optimum deflection decreases
with increasing velocity and decreasing void fraction.

Cyclone I without a hub gave 25% lower pressure drop, but produced a sig­
nificantly thicker air cone more diffused with water. For this reason a hub
seems essential.

The photographs taken during the tests showed that the centrifugal force pro­
ducing separation simultaneously reduces the size of the air bubbles, which
hinders separation. An increase in rotation, hence centrifugal force, above
an optimum value (which varies with the flow conditions) leads to a decrease
in separation. Two reasons were given as to why smaller bubbles hinder
separation.

E-6
1. When large bubbles break up to form smaller bubbles, there is
more surface area; therefore, there should be more drag force on
bubbles retarding separation.

2. For a high gas volume content, the bubbles are not embedded in a
continuum of 1iquid, but the liquid forms thin 1ayers between the
bubbles. Analysis shows that the relative velocity of the liquid
with respect to the bubbles is directly proportional to the square
of the thickness of the liquid layer. Smaller bubbles produce
thinner layers and therefore separation proves to be more diffi­
cult.

The most significant results of these tests is that there is an optimum vane
angle for separation. The cyclone with an inlet angle of 90° and exit angle
of 50° at the hub and 30° at the separator wall appears best for a wide range
of flow conditions. Another significant result is that a hub improves sepa­
ration.

Rouhani has a patent (Reference 3) on a separator with axial vanes (Figure E-4).
He suggested the design guidelines without providing technical justification
which are as follows:

1. Exit angle should be from 20° to 40°.

2. Entrance angle should vary from 90° to 85°.

3. Top edge of vane should be straight and radial with centerline of


pipe.

4. The hub should have a diameter 25-35% of the diameter of the


entrance pipe.

5. Twelve vanes are suggested.

6. The bending radius should increase from the center toward the
periphery.

7. The wall slant should be 13°.

Rouhani1s design has similarities with other axial vane designs reviewed. The
hub should occupy 25-35% of the diameter of the entrance pipe, and the exit
angle should be between 20° and 40°.

Dement'ev in References 4 and 5 presents results of tests intended to optimize


the swirler design (Figure E-5). A series of swirlers differing in number of
vanes, in their angle of inclination, and in the diameter of the hub were

E-7
tested. The following optimum design was determined:

1. 6 vanes.

2. Exit angle with respect to horizontal: 35° at wall and 67° at hub.

3. 15° overlap.

The central gas core retained some water in the form of fine droplets and
broken films. The peripheral annular layer of water contains few air bubbles.
Attempts at removing the moisture from the air core by increasing the centri­
fugal effect or by increasing the height of the vanes was unsuccessful.

There are several results of Dement'ev tests which are significant. Six vanes
with an exit angle of 35° proved optimum. Similar to Traube's tests (Reference
2), Dement'ev observed water in the air core. Dement'ev like Traube observed
decreasing the exit vane angle below 35° did not improve separation.

Cochran discusses the design of an axial flow vapor-liquid separator in


Reference 6. Eight to twelve axial vanes (Figure E-6) are used in this design.
The inlet angle is 90° (vertical), and the exit angle is 30° to 40°.

This axial vane separator 1ike many other axial vane separators has an entrance
angle of 90° and exit angle of 30° to 40°. This type design is presently being
used in boiling water reactors.

The review of the 1iterature revealed similarities among the axial vane designs.
Usually 4 to 8 vanes were used. The inlet angle was about 90° (vertical) and
the outlet angle about 35° at the wall and 50° at the hub. A gentle curvature
from the inlet to the outlet exit angle of the vane is preferred. A hub to
separator diameter ratio of about 25% seems to be a nominal value. Several of
the references indicated that there was excessive amounts of water in the steam
core after separation which indicates a secondary separator (discussed later in
this Appendix) may be needed if axial vanes are used. In Appendix H,
Recommended Tests, the above vane geometry was recommended as the nominal de­
sign for one of the separator concepts for development.

E-8
Figure E-2. One stage of vanes.

- Hub

Vane

Figure E-3. Helical vanes.

E-9
Top Angle

Figure E-4. Rouhani's Axial Vanes.

E-10
Section 3

PRE-DRYERS

A pre-dryer is usually an integral part of the separator located at the exit end.
It performs one or more of the following functions:

1. Extracts water from moisture laden steam,

2. Causes small droplets to coalesce for easier separation, or

3. Distributes the steam over the effective flow area above the
separator.

Literature on some of the more commonly used pre-dryers is summarized: Cochran


discusses a pre-dryer (Figure E-7) in Reference 7. Moisture laden steam travels
in a rotary motion into the space between the mesh cone and mesh cylinder. The
rotary motion causes the mixture to flow towards the mesk cylinder where part of
the rotary motion is eliminated before traveling through the mesh cone where most
of the remainder of the rotary motion is eliminated and more of the moisture
droplets coalesce. The water droplets either drain or fall back to the periphery
surface which acts as a dam. The water drains out through the elongated slots.

Another pre-dryer design (Figure E-8) which has been used in marine and fossil
applications is wire mesh discs contained in a perforated cylinder. A number 7
mesh can be used with alternate layers crimped. The purpose of this pre-dryer
is two fold: it causes some of the water droplets to coalesce and it distributes
the flow over more of the effective flow area above the separator.

Corrugated chevron plates (Figure E-9) are sometimes used as a pre-dryer. In


this case the pre-dryers can be defined as a first stage dryer. Moisture adheres
to the chevrons and drains down the corrugations. Ravese in Reference 12 des­
cribes one such pre-dryer.

Another type pre-dryer (Figure E-10) is the deflector plate (Reference 8). The
main purpose of the deflector plate is to prevent the steam from jetting from

E-12
the separator and thus distributes the flow into a larger area. Provisions are
made in this concept to allow the moisture that adheres to the surface of the
deflector to drain to the outer wall of the separator.

Sokolowski in Reference 9 describes a turbo separator (Figure E-ll) with tan­


gential ports. In Appendix B the tangential ports of Design 2 separator were
discussed. These constitute a type of pre-dryer and consequently will be
further discussed here. These ports probably prevent a secondary vortex from
forming. (If they were not used, part of the water film might reverse direction
at the exit nozzle and be entrained by the steam.) The water that exits through
these tangential ports is slung on the wall of the steam generator and drains
downward. The steam rises to flow through the dryers.

Another pre-dryer (Figure E-12) is described in Reference 10. By the pre-dryer


having a smaller diameter than the main separator and the separator not extending
to the height of the pre-dryer, there is a relatively large open space outside
the separator in the pre-dryer region. Gravity separation can become more
effective in that region. Centrifugal motion exists in the flow when it enters
the pre-dryer. The small diameter of the pre-dryer increases the water vortex
thickness. Therefore, the water vortex is more easily skimmed off.

In Reference 11 a pre-dryer (Figure E-13) is described with secondary vanes.


The main separator (not shown) extends and joins the bottom of the pre-dryer at
the middle cylinder. From the main separator, the dryest steam enters the center
cylinder and exits, whereas, the lower quality mixture enters between the center
and middle cylinder. The steam and water mixture is directed through slots into
the channels between the secondary vanes. The secondary water exits both down­
ward and through slots in the outer jacket, while the steam exits in the annul us
between an intermediate cylinder (located between the middle and outer cylinders)
and middle cylinder.

Guide vanes (Figure E-14) can be used as a pre-dryer. The guide vanes should be
designed to distribute the flow over the flow area in the shortest distance to
make maximum use of this height. Moisture laden steam enters the guide vanes
from the separator (not shown) and is distributed over the area above the sepa­
rator. Jetting from the separator exit nozzle (not shown) is minimized and
maximum use of gravity separation is made.

E-13
There is limited data on the preformance of the previously described pre-dryers.
In Appendix H. Recommended Test Program, the pre-dryers described in this appen­
dix are recommended for test on a comparative basis in an air/water loop. In
these tests, the operation of the pre-dryers could be visually observed and the
performance monitored.

E-14
SI-3

Figure E-7. Mesh Cone and Cylinder.


Dry Steam

fi
i

Wire Disks
Side View

H
iffl-
;nnt -
•—
Water
mg&-
Perforated B*..
Cylinder

■ureT-aflen SteSW
Moisture .

Figure E-8. Screen Assembly.

E-16
partial section "a A*

Steam ^Top Plate

Steam

Corrugated
Water Chevrons

Moisture Laden
Steam

Figure E-9. Corrugated Chevrons.

E-17
Deflector Plate

Steam

Steam

Drain Pipe
E-18

Moisture Laden
Steam

Water

Figure E-10. Deflector Plate.


Figure E-ll. Turbo Separator with Tangential Ports

E-19
-j Steam |—

Skimmer

^ Pre-dryer
Water

Water
/ \

*
I
I f
\
Main Separator

Figure E-12. Second stage skimmer.

E-20
Steam Exit

Center
Cylinder Secondary Vanes

Channels

oo o °
Steam''
Water Slots
Mixture

Middle Cylinder
Steam

Outer Cylinder

Figure E-13. Pre-dryer with secondary vanes.

E-21
Figure E-14. Guide vanes.
Section 4

SECONDARY SEPARATION DESIGN

A secondary separation device is sometimes used in those separators with a down­


comer such as the one shown in Figure E-I. After the steam and water are
separated in a separator, the water forms a cone on the inside of the inner
jacket and the steam migrates to the core of the separator. The water is
usually channeled into the separator downcomer by a skimmer or perforations in
the inner jacket. In many cases, steam is entrained in the exiting water. This
entrained steam (carryunder) can be separated from the water by a secondary
separation device.

The following are summaries of secondary separation devices reviewed in the


1iterature survey.

One secondary separation device is a baffle arrangement described by Robbins in


Reference 10. Figure E-15 is a partial layout of the outer jacket of the
separator. Water with entrained steam is picked-off by the skimmer and channel­
ed into the separator downcomer (only the outer jacket is shown on Figure E-15).
The fluid enters the downcomer with a counterclockwise rotary motion. The
baffles arrest the rotary motion and turn the flow downward. The resulting
centrifugal force causes the steam to flow perpendicular to the water toward the
rear members. The steam flows through the perforations and escapes through the
exit openings. The water flows downward and exits at the bottom of the separator
downcomer (not shown).

Dement'ev used hollow vanes (Figure E-16) in the downcomer annul us to separate
the steam that was carried under (Reference 4). Dement1ev employed this device
in an axial flow separator. After the water and steam pass through the vanes
and are separated, the water forms a vortex on the inside of the inner jacket
and the steam forms a vortex in the center of the separator (not shown). The
water vortex with entrained steam climbs the inner jacket of the separator and
flows through ports into channels between the hollow vanes. The hollow vanes

E-23
maintain the centrifugal motion of the flow. The water is pressed toward the
separator casing, and the steam passes through the slits at the end of the
hollow vanes and flows upward. The water flows out the bottom of the downcomer
annulus between the inner and outer jackets.

One proposed secondary separation device is a "bubble rake" scheme (Figure


E-17). Short tubes cut in half can be used for these rakes. These rakes have
about a 30° downward slant to prevent water from exiting through the openings
in the wall. As the water with entrained steam in the downcomer flows by these
rakes, a vortex is formed. The separated steam generated flows along the rakes
and out the openings.

Axially disposed vanes (Figure E-18), usually located in the bottom portion of
the separator downcomer, have been proposed. The steam and water are separated
by centrifugal motion imparted to the flow. The residual steam forms a vortex
at the end of the short leaf of the vanes. It is exhausted through the exit
port. The steam then joins with the essentially dry steam from the separator
exit nozzle.

Cochran in Reference 6 used a combination of flow channels and vanes (Figure


E-19) for secondary separation. The steam/water mixture is picked-off by the
inner skimmer. The mixture flows horizontally through the mixture flow channels
and down the outer annulus between pairs of vanes. The water exits down the
annulus and the steam upward between two vane pairs.

Most of the previously described secondary separation devices are recommended


for tests in air/water. (The Recommended Test Program is contained in
Appendix H.) The purpose of these tests is to observe the operation and moni­
tor the performance. Effectiveness of devices in separating entrained steam
should be evaluated in air/water tests before incorporation into separator
designs. The most effective secondary separation device should be incorporated
into a separator design and tested further using air/water.
Skimmer

Steam Water Mixture

Members

Perforations

Exit Opening

Steam
Downcomer Flow
Water
Flow

Baffles

Figure E-15. Baffle Arrangement.

E-25
Vane Channel Downcomer

Figure E-16. Hollow Vanes.

E-26
"'Water with Entrained Steam

Downcomer

Bubble Openings
Rake

Steam

Water

Bubble Rake

Figure E-17. Bubble Rakes.

E-27
Exit Port
Axially Disposed
Water with Entrained Steam Vane
Bottom
Portion
Separator
Downcomer
Steam
E-28

I
Water

Figure E-18. Axially Disposed Vanes.


MIXTURE FLOW
CHANNEL

STEAM ESCAPE
> CHANNEL

MIXTURE FLOW
" CHANNEL

SECTION "A - A

VANES

SECTION 11B - B
Inner Skimmer Ring

Mixture Flow
Channel

Outer
Annul us

Figure E-19. Flow Channels and Vanes.

E-29
Section 5

AUXILIARY PARTS TO PREVENT REENTRAINMENT

After the steam and water has been separated by the vanes, the water forms a
vortex on the inside of the inner jacket and the steam migrates to the center
of the separator. At some point along the jacket, the water has to be channel­
ed away from the separator with minimum steam reentrainment. Several separation
designs are used to prevent this. The following are summaries of some of these
designs.

The General Electric and Kraftwerk Union separators shown on Figure E-20 use
skimmers. Referring to the General Electric design (Figure B-10), the water
vortex on the inner jackets is redirected into the annuli by the skimmers. A
restrictor located near the bottom of the lower annulus throttles the water
establishing a water vortex thickness. This thickness is greater than the
clearance of the first skimmer. The first skimmer only picks-off water, the
second and third skimmers pick-off water and/or steam according to the inlet
water flow. As a result of using three skimmers, the water velocities in the
second and third skimmers are lower. More entrained steam bubbles are allowed
to escape from the water to rejoin the steam. The multiple skimmer arrangement
allows a wide range of flow rates with minimum steam reentrainment. The lower
velocities result in less reentrainment than separator designs with a single
skimmer.

Combustion Engineering produces a centrifugal vane separator with a single per­


forated jacket (Figure B-31). The water vortex on the inside of the jacket
discharges the water through the perforations to the separator support deck.
The larger perforations at the bottom of the separator discharge the major
portion of the water and the upper perforations the remainder of the water.
Steam reentrainment can be reduced by raising the water level around the
separator so that the exiting water goes into a water environment rather than
a steam environment.

E-30
One proposed design to channel the water away from the separator with minimum
steam reentrainment is shown in Figure E-21. In this design an "artificial"
water level is maintained around the separator with a water seal. The major
portion of the water discharges through the lower perforations to the
"artificial" water pool maintained by the water seal. The water then flows
over the water seal into the separator downcomer and exits at the bottom of
the separator downcomer. The "artificial" water level maintained by the water
seal allows the water to exhaust into a water environment preventing mixture
with steam which could result in large amounts of reentrainment.

Another proposed design to channel the water away from the separator with
minimum reentrainment of steam is shown in Figure E-22. After separation, the
water flows through the slots in the inner jacket into the separator downcomer,
and the steam exhausts through perforations in the separator cone. The reverse
scoops in the downcomer are located about the slots. The reverse scoops re­
direct the water into the downcomer and prevent it from deflecting off the
outer jacket wall. This gradual change of direction for the water prevents it
from interacting with the steam in that portion of the separator downcomer,
thus minimizing steam reentrainment.

In many cases, vanes do a good job of separating the steam and the water. If
the water is not properly channeled away from the separator, then steam can be
reentrainment. In Appendix H, Recommended Test Program, a separator with two
skimmers and a restrictor is suggested for testing. The purpose of these tests
is to optimize the restrictor and skimmer clearances for the flow rate range of
this separator.

E-31
Design 3 Separator Design 4 Separator
(General Electric) (Kraftwerk Union)

Skimmers

Restrictors

Figure E-20. Separators with Restrictors and Pick-Off Rings.

E-32
Downcomer
N
Perforated Wall

Perforated Wall^ ,
-Water Seal
Watef

Downcomer

Steam and Water

Figure E-21. Water Seal Separator.

E-33
Steam Flow

Slots

Reverse Scoop

Helical Vanes

Downcomer

Water

Steam and Water Flow

Figure E-22. Separator with Reverse Scoops.

E-34
REFERENCES

1. "Primary and Secondary Separation of Steam and Water by Mechanical Means."


Allis Chalmers Nuclear Power Report 65003, 1965.

2. K. Traube and J. Vollradt. "Investigation of the Separation of Flowing


Gas/Liquid Mixtures by Centrifugal Forces." Chemical Engineering
Technology, No. 12, Vol. 14, 1962, pp. 827-833.

3. S. Z. Rouhani. "Separator for Mixture of Mixture of Steam and Water."


Swedish Patent 373451 , 1975.

4. B. A. Dement'ev, et al. "Investigation of a Separator Unit with Axial


Supply of Gas/Liquid Mixture." Teploenertika, No. 3, Vol. 19, 1972,
pp. 25-29.

5. B. A. Dement'ev and Y. I. Maiinin. "Experimental Investigation of


Centrifugal Steam/Water Separator. Teploenertika, No. 9, Vol. 21, 1972,
pp. 78-79.

6. J. T. Cochran, H. D. Ongman, and H. E. Weber. "Axial Flow Vapor-Liquid


Separator." U. S. Patent 3216182, 1965.

7. J. T. Cochran, "Pressure Recovery Axial Flow Vapor-Liquid Separator."


U. S. Patent 3329130, 1967.

8. R. Bouecke and G. Schucktanz. "Two Phase Flow Separation in U-Tube


Steam Generators." Proceedings of 17th National Heat Transfer Conference,
1977.

9. P. F. Sokolowski. "Vertical Steam Drum." U. S. Patent 3751886, 1973.

10. C. H. Robbins, et al. "Gas Liquid Separator." British Patent 1259711 ,


1969. '

11. Steam Separator in BWR. Japanese Patent 41-18365.

12. T. Ravese. "The Application and Development of the Turbo Steam


Separator." Combustion Engineering, 1954.

E-35
APPENDIX F
RECOMMENDED SEPARATOR CONCEPTS
ILLUSTRATIONS

Figure Page

F-l KWU Compact Separator F-4


F-2 Carryunder Versus Water Level for KWU Compact F-5
Separator
F-3 Moisture behind Separators Versus Water Flow F-6
Rate
F-4 Elevation View of Up-flow Down-flow Separator F-7
F-5 Plan View of Up-flow Down-flow Separator at F-8
Elevation A-A.
F-6 Summary of Vane Design F-9
F-7 Centrifugal Separator F-12
F-8 Carryover Performance F-13
F-9 Stacked Arrangement in Parallel F-14
F-10 Castellation and Si it Modifications F-15
F-ll Summary of Flow Loadings F-18
F-12 Composite Axial Vane Separator F-19

F-i
APPENDIX F

RECOMMENDED SEPARATOR CONCEPTS

During this study program, the operation and performance of individual design fea­
tures as well as complete separators were evaluated. The design features and sepa­
rators evaluated were of proposed and presently used separation equipment. As a
result of this study, the following general types of separators were recommended
for further development:

1. Centrifugal Vane, Upflow, Downflow Separator

2. Centrifugal Vane, Perforated Jacket Separator

3. Axial Flow Separator

Three criteria were used for the selection of the separators recommended for further
development. First, the separator should have potential of operating under high
steam and water flow loadings while maintaining a low pressure loss, because of the
desirability of increasing the circulation ratio in contemporary nuclear steam
generators. This would necessitate redesigned separators in order to maintain allow­
able carryover and carryunder. Second, the separator should show promise of having
very low carryover (perhaps less than 0.1%) over a wide range of steam and water
flow conditions and water levels. Carryover can result in transport of solids or
corrosive impurities into the turbine cycle. Carryunder should not be adversely
affected to avoid significant degradation of recirculation ratio in most steam
generator designs. It should be less than 1%, preferably in the 0.6% range or less.
Third, the separator should be compact enough to be retrofitted to existing steam
generators. The separator diameter should be in the range of 12" so that it can be
passed through manways of existing steam generators, and its height should be less
than 60" so that it can fit into the available drum space of most nuclear steam
generators. One or more of these three criteria were used to select each of the
three general types of separators.

Following is a discussion of each of the three separators recommended for further


development. Each of these separators are specific designs and are not necessarily
the only model produced by a manufacter. Their operation (steam/water flow paths)
was discussed in Appendix B.

F-l
CENTRIFUGAL VANE, UP-FLOW, DOWN-FLOW SEPARATORS

The centrifugal vane, up-flow, down-flow separator is recommended for further


development because it has the potential of satisfying two of the forementioned
criteria: Capacity for high steam flow rate and compact size to be retrofitted
into existing steam generators.

From the review of the 1iterature, it is apparent that the forerunner of the standard
Kraftwerk Union down-flow, up-flow separator (described in Appendix B) is a separator
tested by Brandes (Reference 2) and shown in Figure F-4 and F-5. Kraftwerk Union
developed and tested a variation of their standard up-flow, down-flow separator
which they called the compact separator (Figure F-l, Reference 1). The allowable
steam flow rate for the compact separator on an area basis is among the highest
of al1 the separators evaluated (See Comparison Table, Figure F-ll). The high steam
loading of the compact separator (Reference 1) shows that there is potential for a
centrifugal vane, up-flow, down-flow separator to satisfy the first criteria of
high steam loading. The separator recommended for further development should be
similar to Brandes1s separator, but with some of the design characteristics of the
compact separation.

A study of the plots on Figures F-2 and F-3 (Reference 1) revealed the compact
separator has a very high allowable steam flow rate, but for a 1imited band of
water flow rates and water levels. Figure F-2 shows that for low water levels,
carryunder increases sharply, and Figure F-3 shows that for high water flow rates,
the exit moisture from the separator increases exponentially. The annular restrictor
clearance at the recirculating water outlet near the bottom of the separator is
a critical design dimension, and is a parameter in Figures F-2 and F-3. As dis­
cussed in detail in Appendix B, the restrictor clearance controls the band of
water flow rates and water levels for which this separator operates with adequate
performance.

To obtain a wider range of water flow rates and water levels, modifications of the
Brandes up-flow, down-flow separator shown in Figures F-4 and F-5 are recommended
for further development. With the up-flow path of the compact separator
(Figure F-l), not closed by a top plate as is the separator shown in Figure F-4,
the water and steam flow have an extra escape direction, resulting in lower velo­
cities. There should be less reentrainment after separation in this type concept.

F-2
The dimensions of various centrifugal vane separators (#4, #5, #6, and #7) are
1 isted in Figure F-6. Brandes (Reference 2) tested various centrifugal vane,
up-flow, down-flow separator configurations. Of the various configurations tested
by Brandes, Separator No. 1 in Figure F-6 most nearly matched the dimensions of the
other four centrifugal vane separators reviewed and in commercial use. A similitude
analysis developed by Brandes (Reference 2) was used to scale the No. 1 base design
separator down to 13.8 inches outside diameter so that it could be passed through
the manways of most nuclear steam generators. Two separators designs (No. 2 and
3 in Figure F-6) resulted from the similitude analysis. Separator design No. 2 is
recommended for further development because the vane exit angle ip was more near the
exit angles of the other centrifugal vane separators evaluated. By developing No. 2
Separator in Figure F-6, the criteria of the separator being smal1 enough to retro­
fitted into existing steam generators should be satisfied.

In summary,a variation of the centrifugal vane up-flow, down-flow separator


(Figures F-4 and F-5) is recommended for further development. The goals of this
development would be to develop a separator which has a high allowable steam flow
rate and is compact in size. Furthermore, this separator concept has the potential,
through development, to perform acceptably over a wide range of water levels.

F-3
Deflector
Plate

/ ///— Drain
/,{ Pipes

Upper
Riser

Vanes

_ Water
Downcomer

Lower Riser
— Steam

Restrictor —Water
Seal

Base
Plate

Figure F-l. KWU Compact Separator.

F-4
Annular Restrictor
1,5- Clearance Gap

-—*—S=0-80 inches
\
\ -------- S = 0.60 inches
\
S =0.46 inrhpc

1.0- Ws = 84000 Lbm/hr


Ww = 297000 Lbm/hr
_gS-
i

0,5-

0 I i
•66 1.31 1.97 2.62 3.28 3.94 4.59

h-[ Water Level (ft)

Figure F-2. Carryunder Versus Water Level for KWU Compact


Separator.

F-5
Moisture From The Separator

Annular Restrictor
Clearance Gap
S = 0.45 inches

S = 0.60 inches

h- 3.28 ft

83800 Lbm/hr

176.4 352.8 529.2

Water Flow Rate x 10 Lbm/hr

Figure F-3. Moisture behind Separators Versus Water Flow Rate.


Deflector
Plate

Downcomer

Upper
Riser

Top Plate

Lower
Riser

Water Steam
Base Plate

Figure F-4. Elevation View of Up-flow Down-flow Separator.

F-7
Figure F-5. Plan View of Up-flow Down-flow Separator at
Elevation A-A.

F-8
df ^s hS ¥ Vane Spacing at Entrance

O
Brandes, Base Design 16.1 13.8 2.75 76° 32°

O— i
0

oo
Brandes 13.8 11.8 3.15 76 25°

r*~i
Similitude, Shock loss

Brandes 13.8 11.8 3.15 68° 27° 10°


Similitude, Separation

oo
Combustion Engineering, 11.0 8.0 3.5 48° 15°

CO
Inc.

Kraftwerk Union Compact 19.3 13.4 5.9 81° 25° 10°

o
Kraftwerk Union 15.5 2.4 20°

CO
12.6 10°

Curtis Wright 15.3 8.0 4.0 91° 20o-28° 15°

Figure F-6, Summary of Vane Design.


CENTRIFUGAL VANE, PERFORATED JACKET SEPARATOR

The centrifugal vane, perforated jacket separator is recommended for further


development because it has the potential of satisfying two of the three criteria
set forth for selection of separators. The first criterion should be satisfied is
that carryover is very low over a wide range of flow conditions and water levels.
The second criterion that should be satisfied is that this separator is compact
enough to be retrofitted into existing steam generators.

Figure F-8 illustrates the performance of the Combustion Engineering, Inc. centri­
fugal vane, perforated jacket separator shown in Figure F-7. At nominal water
level, carryover is 0.5% for a steam flow rate of 60000 Lbm/hr. It decreases to
0.05% for a steam flow rate of 45000 Lbm/hr, and decreases further at 0.01% for
a steam flow rate of 35000 Lbm/hr. At a steam flow rate of 35000 Lbm/hr, carryover
remains in the negligible range up to and beyond water levels two times the nominal
value. At low steam flow rates, this separator meets the criterion of having low
carryover from the separator. Since this separator has outside diameter less than
12 inches, it meets the third criterion of being able to be retrofitted into
existing steam generators.

Two modifications of this concept are recommended for further development. One
modification as described in Appendix D could be to stack the separators in
parallel (Figures D-l and F-9). A second modification could be to replace the
perforations with castellations or slots as shown in Figure F-10.

The centrifugal vane separator in the stacked arrangement (Figure F-9) operates
in the following manner. The steam and water rises from the tube bundle of the
steam generator to enter the standpipes of the lower and upper separators
(Figure F-9), The mixture is redirected by a top plate through centrifugal vanes.
The water is slung against the inner wall, flows through perforations in the inner
wall into the downcomer and exits downward onto the separator deck. The steam mi­
grates to the center of the separators and exits through the exit nozzle at the top
of the separator.

By stacking the centrifugal vane, perforated jacket separator in parallel as shown


in Figure F-9, the amount of steam flow per separator is decreased in comparison
with conventional arrangements, and there is a potential that carryover at the
separator exit will be negligible. As shown in Figure D-l and discussed in
Appendix D, one stacking arrangement would be to space the 12" outside diameter

F-10
separators on a 14" square pitch instead of the conventional triangular pitch
arrangement. This creates a 33% increase in the number of separators and a 25%
reduction in flow rate to each separator, thus decreasing pressure drop and im­
proving carryover performance.

A second modification to the centrifugal vane, perforated jacket separator could


be to replace the perforations with castellations (Figure F-lOa) or slots
(Figure F-lOb). Dement'ev in Reference 3 indicated that this type of modification
was superior to perforations and should improve carryover performance.

In the operation of the centrifugal vane separator, the water rotates in a spiral
direction upward along the inside of the separator. The water exits radially
through the perforations. The centrifugal vane separator with the slot modification
would provide a tangential entrance of the separated water into the downcomer.
This modification should minimize interaction of the separated water with steam
outside the separator barrel and thus reduce carryunder.

In summary, further development of modifications to the centrifugal vane, perforated


jacket separator may result in separators with negligible carryover. This may
eliminate the need for dryers in nuclear steam generators. Furthermore, this type of
separator is usually compact in size, making it potentially useful in retrofitting
to existing steam generators.

F-ll
Exit Nozzle

Perforations

Centrifugal Vanes

Inlet Nozzle

Figure F-7. Centrifugal Separator.

F-12
A
Carr
%

Water Level (l=nominal value)

Figure F-8. Carryover Performance

F-13
Exit Nozzle

(Not to Scale)

Perforated Jacket

Downcomer

Vanes

Standpipe

Support Deck

Water Steam

Figure F-9. Stacked Arrangement in Parallel.

F-14
Steam
Steam I

Castel!ations
Slits

Water

•Downcomer

Vanes

Water

CUTAWAY

a. CASTELLATIONS MODIFICATION b. SLIT MODIFICATION

Figure F-10. Castel1ation and Slit Modifications.

F-15
AXIAL FLOW SEPARATOR

Two criteria were used in selecting the axial flow separator for further development:
a high allowable total flow rate and being compact enough to be retrofitted into
existing steam generators.

The flow loading of the General Electric axial flow separator (Reference 4) was com­
pared with eight other separators (Summary of Flow Loadings, Figure F-ll). The flow
loading used was the flow rate at which 0.25% carryover occurred from the respective
dryer systems. The General Electric axial flow had the highest allowable total
flow loading on an area basis, and the second highest allowable steam flow loading
on an area basis. Because of the relatively high allowable total flow loading,
there is a strong potential that the axial flow separator will best satisfy the
first criterion of allowing high steam and water flow loadings.

The performance and operation of the General Electric axial flow separator is
discussed in Appendix B.

The General Electric axial flow separator (Figure B-10) is approximately 12 inches
in outside diameter and 90 inches in height. The diameter of this separator is
smal1 enough to fit through the 16 inch manways of most steam generators, but the
90 inch height is longer than the 60" desired drum length of most nuclear steam
generators. Since, as discussed in Appendix B, the third skimmer channels water
at high flow conditions only, a few design modifications might be made to reduce
the height of this separator to the 60 inch and thus satisfy the height criterion
for retrofitting into existing steam generators. The design modifications re­
commended for further development are to eliminate the third skimmer and calibrate
the restrictor, which controls water flow to the skimmers, for two skimmers instead
of three.

A schematic drawing of a possible composite axial flow separator recommended for


further development is shown in Figure F-12. This axial flow separator has the
same major design features as the General Electric axial flow separator and in­
cludes some of the modified design features discussed herein. The third skimmer
has been eliminated and may possibly be replaced by a screen assembly. The
skimmers are round edged instead of square edged. There is inlet transition area
before the vanes. The function of these design features are discussed in Appendix H.
If developed further, a composite separator should be designed around optimizing
each design feature.

F-16
In summary, the axial flow separator is recommended for further development in
order to have a compact separator, less than 60 inches in height and about 12 inches
in diameter, and in order to have a high total allowable flow rate, possibly as
much as twice the allowable flow rates for the other separators recommended in this
appendix.
»3
Wt/A X 10“3 Lbm/hr Ws/A X 10 Lbm/hr

General Electric 467 60


(Reference 1)

Russian 331 47
(Reference 2)

Combustion Engineering 147 37

Curtis Wright 91 15
(Reference 3)

Westinghouse 163 40
(Reference 4)

Kraftwerk Union 115 33


(Reference 5)

Radial Vane 122 12


(Reference 6)

Kraftwerk Union 242 67


(Reference 5)

AEG Downflow 168 21


(Reference 7)

Figure F-l1. Summary of Flow Loadings.

F-18
A
I

Screen Assembly

—- Upper Skimmer

Straightening Vanes

Lower Skimmer

Straightening Vanes

Restrictor

Axial Vanes

Inlet Transition

Support Plate

Figure F-12. Composite Axial Vane Separator.

F-19
REFERENCES

1. R. Bouecke and G. Schucktant. "Two Phase Flow Separation in U-Tube


Steam Generators." In Proceedings of 17th National Heat Transfer
Conference, 1977.

2. H. Brandes, "Experimental Investigation of Centrifugal Separators -1."


Energie and Technik, Vol. 23, 1971, pp. 283-290.

3. B. A. Dementier and et al. "Experimental Investigation of Centrifugal


Steam/Water Separator." Teploenergetika, Vol. 19, No. 3, 1972.

4. R. H. Moen and S. Wolf. "Advance in Steam-Water Separators for Boiling


Water Reactors." ASME 73-WA/PWR-4, 1973.

F-20
appendix g

PROMISING SEPARATOR CONCEPTS


ILLUSTRATIONS

Figure Page
G-l Separator with Hollow Vanes in Annulus G-2
G-2 Performance of Separator with Hollow Vanesin Annulus G-3
G-3 Separator with Converging Shell G-7
G-4 First Generation Modified Separator G-10
G-5 First Generation Separator G-ll
G-6 Carryunder Versus Inlet of Quality G-12
G-7 Carryunder Versus Water Level G-12
G-8 Carryover at Dryer Inlet G-12
G-9 Carryover from Dryers G-12
G-10 Second Generation Separator G-13
G-ll Axial FIow Venturi Separator G-14
G-12 Performance Versus Inlet Quality G-15
G-13 Performance Versus Water Level G-15
G-14 Radial Vane Separator G-18
G-15 Vane Pair G-19
G-16 Typical Carryover and Carryunder Curves asa Function G-20
of Water Level in FCRV-2
G-17 Carryover Before Break Point versus Homogeneous Nozzle Velocity G-21
G-18 Influence of Nozzle Velocity on Separator G-22
G-19 Turbo Separator G-24
G-20 Downflow Separator G-27
G-21 AEG Cyclones G-31
G-22 AEG Separator with Four Outtakes G-32
G-23 Allis Chalmers Downflow Separator G-33
G-24 Performance Curves for Separator of FigureG-21A G-34
G-25 Performance Curves for Separator of FigureG-22 G-35
G-26 Air water test of Allis Chalmers Separator G-36
G-27 Performance at 600 psi G-37
G-28 Performance at 1200 psi G-38
G-29 Performance at 1800 psi G-39
G-30 Efficiencies of Axial Flow Venturi Separator G-42

G-i
APPENDIX G

PROMISING SEPARATOR CONCEPTS

During this program, separators which are presently in use for various applications
and those proposed for use were evaluated. Three separators which have potential
for satisfying one or more of the foil owing criteria were recommended for further
development and are described in Appendix F. The criteria are:

1. Can be retrofitted into existing nuclear steam generators.

2. Has potential for low carryover.

3. Has potential for high allowable flow loadings.

Obviously, other separators also showed promise. This Appendix presents a detailed
discussion of ten promising separator concepts. This discussion covers reasons
as to why they are considered promising, the operation of each concept, the per­
formance of each concept where test data is available, and reasons as to why these
separators are not recommended for further development. In choosing the ten pro­
mising separator concepts, priority was given to those separators which are pre­
sently in use in boiling water reactors, nuclear steam generators, and boiler drums.
Also, priority was given to those separators which have been studied extensively.
The manufacturer or origin of each separator is given for each design evaluated.
The reader is cautioned that manufacturers may produce other model separators.

G-iii
p
I
Section 1

SEPARATOR WITH HOLLOW VANES IN ANNULUS (Russian)

A separator with hollow vanes in the annul us is shown in Figure G-l. The steam
and water mixture enters the separator pipe and flows through the vanes. The
separated water flows upward along the inside surface of the separator pipe. The
water with some entrained steam exits through slots (overflow ports) in the
separator pipe. The water with the entrained steam then flows between the hollow
vanes in the annul us. As a result of centrifugal force, the water is si ung
against the separator casing and exits downward through the annul us. The
entrained steam flows through the gap at the end of the hollow vane into the
hollow vane and rises to exit the separator. The main portion of the steam rises
through the separator core to leave the separator.

The carryover performance of this separator for operating pressures of 160, 305
and 522 psi is correlated as a function of kinetic flow energy and is shown in
Figure G-2 (Reference 1). For a steam flow rate of 21400 Lbm/hr, water flow
rate of 48400 Lbm/hr, and operating pressure of 510 psi, the kinetic flow energy
o
is 255 Lb/ft . Using Figure G-2, the carryover from this separacor at these
conditions is 9.5%.

The diameter of the separator casing is 7.61" and the separator barrel diameter
is 3.9". Six vanes are attached to a 1" diameter hub. The overflow ports are
27.6" high and located 35.4" above the vanes. The total separator height is
slightly greater than 63".

This separator was not recommended for short term development because:
1. No test data is available in the pressure range typical of US PWR
steam generators.
2. High carryunder in the high flow range.
3. Difficulty in fabricating secondary vanes.
4. 10% carryover in high flow range.

G-l
!. PRESSURE TAP: 2. FLANGE; I. SEPARATOR PIPE 2. SWIRLER
3. SWIRLER: 4. SEPARATOR PIPE;
3. CENTRAL SPINDLE 4. VANE
5. BOX. SQUARE IN CROSS SECTION;
6. CASING OF SEPARATOR: 7. VANE
8. COVER

I. CASING OP THE SEPARATOR; 1 SEPARATOR PIPE;


3. OVERFLOW PORT: 4. VANE CHANNEL; 5. VANE

HOLLOW VANES IN WATER ANNULUS

Figure G-l. Separator with Hollow Vanes in Annulus.

G-2
a 155 psi
□ 300 psi
o 510 j)Si

61 102 143 184 225 266 307

«*m!|V28> LVft2

KINETIC FLOW ENERGY

Figure G-2. Performance of Separator with Hollow Vanes in Annulus.

G-3
Section 2

CURVED ARM SEPARATOR (Curtis Wright)

A description of the operation of this separator is given in Appendix B of this


report under Design 1 Separator. Top and side sketches of this separator are shown
in Figure B-l and Figure B-2. Performance curves are given in Figures B- 3, B- 4 ,
B- 5, B- .6* and B-.6 . (Reference 3)

The carryover performance for this separator (Figure B-7 ) shows that massive carry­
over occurs when the steam flow rate exceeds 22000 Lbm/hr for a circulation ratio of
6. The results were at a pressure of 345 pais. At higher pressures, less separa­
tion and higher carryover would be expected because the density difference between
the steam and water would be less. The curved arm separator is not recommended
because other separators (Appendix F) can accomodate a higher steam flow and total
flow loadings with lower carryover.
Section 3

TURBO SEPARATOR WITH TANGENTIAL PORTS (Westinghouse)

The operation and performance of the turbo separator with tangential ports is
discussed in Appendix B as the Design 2 separator (Reference 4 and Reference 5).
A top view and a side view of this separator are shown in Figure B-9 and Figure
B-8. This concept has the advantage of having the lowest pressure drop of the
candidates evaluated, but the exit moisture was in the 25% range. Excellent
dryers are required to obtain carryover to the turbine of less than 0.25%. The
diameter of this separator is 64 inches which is too large to be passed through
existing manways and hence would prohibit a retrofit operation. Because of this
separator's size and substandard carryover performance, it is not recommended for
further development. No performance curves were available.

6-5
Section 4

SEPARATOR WITH CONVERGING SHELL (Foster Wheeler)

This separator is a patent of Foster Wheeler (Reference 12). The steam and water
mixture enter the lower riser and flow through the primary vanes which impart
centrifugal force to the mixture. The water is slung against the inside of the
inner jacket, and the steam flows to the core. Most of the steam passes through
the upper riser out the top. Because of the taper of the inner jacket, the water
is prevented from flowing over the top. The water passes down the inner passage
with substantial amounts of entrained steam. Secondary vanes at the bottom of this
passage separate the entrained steam. The separated steam rises and the water
flows out the bottom of the downcomer.

This separator was included as a promising concept because a variation to it might


minimize carryunder. This variation would be to increase the angle of the inner
jacket to allow part of the water to flow up and over the inner jacket and part
to descend through the inner passage. By decreasing the water velocity more en­
trained steam bubbles can escape.

A complete understanding of this separator's capabilities was not possible as no


test performance data was available. Therefore no recommendation regarding fur­
ther development is possible.

G-6
SEPARATOR WITH CONVERGING SHELL

Steam
Steam ^ steam
TZ2ZZZL

Upper Riser

Downflow
Water with Entrained Steam

Inner Passage

Primary Vanesi ^

Secondary Vanes

Top View

Lower Riser

Side View

Water Water
_ _ J___

Steam and Water Mixture

Figure G-3. Separator with Converging Shell.

G-7
Section 5

EARLY AXIAL FLOW SEPARATORS (General Electric)

General Electric Company developed the separator they presently use in their
boiling water reactor vessel through a series of testing and experimental pro­
grams.

One of the first separators they used, termed First Generation Modified Separator,
is shown on Figure G-4 (Reference 7). This separator operates similar to other
axial flow separators. The swirler imparts the centrifugal force to separate the
steam and the water. Moisture laden steam from the center of the separator flows
through the exit nozzle into the pre-dryer which extracts some of the excess mois­
ture (pre-dryer described in Appendix E-2). A large portion of the water is picked-
off by the inner skimmer ring and flows through straightening vanes in the downcomer
to exit the separator. The steam and water interface is picked-off by the outer
skimmer ring and flows through a wire mesh.

With several design improvements, the First Generation Modified Separator evolved
into the First Generation Separator (Figure G-5). The wire mesh was replaced with
a more sophisticated secondary separation device (described in Appendix E-3). The
separator diameter was also reduced. The straightening vanes were replaced with
diffuser castings. A restrictor was placed in the bottom of the inner downcomer
to regulate the water film thickness on the inside barrel to insure the primary
skimmer ring would skim water (Reference 8).

The performance curves for this separator are shown on Figures G-6, G-7, G-8, and
G-9 (Reference 6). Figures G-6 and G-7 indicate that for a total inlet flow rate
of 450,000 Lbm/hr, carryunder is in the 0.8% range. Figures G-8 shows that for
water levels less than 45 inches, carryover at the dryer inlet is less than 10%.
Figure G-9 reveals for steam flow rates less than 32000 Lbm/hr, carryover from
the dryers is less than 0.02%.
■s'
Further refinements to the First Generation Separator resulted in the Second
Generation Separator shown in Figure G-10 (References 9 and 10). The wire mesh

G-8
pre-dryer was replaced with a second stage skimmer (Explained in Appendix E-2).
The secondary separation device was modified to a baffle arrangement. Figure E-15
(described in Appendix E-3). The diffuser casting was replaced by a series of
vertical baffles. The diameter was reduced from 13.25" to 12.75". At water level
of 25" and steam flow rate of 470,000 Lbm/hr, Figure G-12 shows carryunder steadily
increasing at an inlet quality of 14% and greater. Carryover exponentially in­
creases at inlet qualities below 10% and gradually increases for inlet qualities
greater than 16%. For an inlet quality of 14% and total flow rate of 470,000
Lbm/hr, Figure G-13 shows carryunder remains in the .2 to .1% range and carryover
exponentially increases when the water level exceeds 40 inches (Reference 11).

Additional improvements were made to the Second Generation Separator to arrive at


the presently used Third Generation Separator. The diameter has further reduced
adding three stages of skimmers in series rather than parallel (Figure B-10 in
Appendix B). The performance and operation of the Third Generation Separator are
discussed in Appendix B as Design 3 separator.

The early axial flow General Electric separators were considered promising because
they employed several design features which are being successfully used in other
separator concepts. Appendix F. For example, Combustion Engineering, in their
Design 2 Separator (Appendix F) uses a pre-dryer which is composed of wire mesh
similar to the General Electric First Generation Separator. Kraftwerk Union in
their Design 1 Separator (Appendix F) uses successive skimmers, restrictor, and
straightening vanes similar to the General Electric Second Generation Separators.
The General Electric Design 3 Separator (Appendix F) employs many of the design
features employed in their axial flow separators such as axial vanes, successive
skimmers, restrictor, and straightening vanes. The early General Electric axial
flow separators are not recommended for further development because they represent
early stages of development of the General Electric Design 3 Separator (Appendix F)
which is recommended for further development.

G-9
Pre-Dryer

...-Exit Nozzle

Pick-Off Rings 'Straightening Vanes

Wire Mesh

Downcomer

Water Steam

Figure G-4. First Generation Modified Separator.

G-10
4
I

Exit Nozzle

Restrictor

Figure G-5. First Generation Separator.

G-ll
WLET X « 5.8 ptfcwt
'total ao» * mm *Mwt***

VEtCNT PERCENT
“CAPSYUNOER
MLETX*7.9percMt
TOTAL FLO# « 440,000

INLET X a
.TOTAL FLO# *

WATgft LEVSL (incftw)

& 4S

SEPARATOR mLETQUAUn (wt%)


SATER LEVEL (tnchM)
Variation of carryunder with inlet quality
(•for total inlet flow of 450,000 Ib/hr/separator) Carryunder versus water level

Figure G-6. Carryunder Versus Inlet Figure G-7. Carryunder Versus


of Quality. Water Level.

SEPARATOR
INLET QUALITY.

LIQUID CARRYOVER IN
VESSEL OUTLET STEAM,
weight percent

- SEPARATOR CARRYOVER.
__ weigM pefcwit «

1 TT I

REACTOR WATER LEVEL (SEPARATOR SKIRT SUBMERSION), inches PRIMARY STEAM FLOW/SEPARATOR, THOUSANDS OF Lfr'Hf

Separator carryover at dryer inlet Outlet steam carryover versus steam flow
rate

Figure G-8. Carryover at Dryer Figure G-9. Carryover from


Inlet. Dryers.

Performance Curves for First Generation Separator.

G-12
*

Pre-Dryer

Secondary Separation
Device

Vertical Baffles

Restrictor

Water Steam

Figure 6-10. Second Generation Separator.


*
i

WATER POOL
WATER DISCHARGE
PASSAGE
---------^ Water

DIVERGING SECTION OF -------------- >.


Steam
VORTEX TUBE

VORTEX FINDER-"'..

THROAT--------^
CONVERGING SECTION OF
VORTEX
GUIDE VANES-"~.. ....... -

INLET SECTION

Figure G-ll. Axial Flow Venturi Separator.

G-14
TOTAL FLOW RATE = 470,000 1
SEP. WATER LEVEL= 25 inches

%
CARRYOVER, WEIGHT
CARR

INLET QUALITY,PERCENT

Figure G-12. Performance Versus Inlet Quality.

1.0

%
CARRYOVER, WEIGHT
0.8

Di
0.6
LU

0.4

0.2

-20 10 20 30 40 50

SEPARATOR WATER LEVEL, inches

Figure G-13. Performance Versus Water Level.

G-15
Section 6

RADIAL VANE SEPARATOR (Atomic Energy Commission)

The development of the radial vane separator (Figure G-14) was sponsored by the
AEC and EURATOM, and the work was conducted by General Electric. The radial
vane separator was designed to be installed over the reactor core of a 300 MWt
Boiling Water Reactor. Scaled down radial vane separators and vane pairs
(Figure G-15) were tested using air/water and high pressure steam/water. Ex­
tensive analysis of the separation process was used in coordination with the
testing (Reference 13, 14, 15, 16, and 17). Figures G-16, G-17, and G-18 show
some of the data collected in these tests.

A radial vane model separator is shown in Figure G-14. Steam/water mixture from
the reactor core enters a plenum bounded above by a top plate and by the shroud
wall on the sides. Guide vanes (not shown) in this plenum turn the mixture from
axially upward to radially outward into the nozzle leading to the passages between
the vanes. In the vanes, the steam moves out from the water due to the effect
of the centrifugal force resulting from flow on the curved surface of the vane.

Separated steam rises in the vane column between the water surface and the next
vane. The water and some steam which is entrained flows off the ends of the vanes
onto the circular reactor wall where it forms a vortex of water accelerating
downward. When the water vortex contacts the water reservoir, a rotating eddy
is formed. This eddy tends to move the carryunder bubbles radially inward where
they are moved upward to the surface and released. The carryunder steam flow
joins the main steam flow and flows out the top of the separator. The water is
recirculated.

Steam and water tests at 1000 psig were conducted on a radial vane model separator
termed FCRV-2 (Figure G-15). For a total flow rate of 450,000 Lbm/hr and inlet
quality of 10% (Figure G-16), carryunder decreased from 1.2% at 1 inch of water
level to 0.3% at 5 inches water level and to negligible amounts at water levels
greater than 25 inches. On the other hand, carryover is less than 0.1% at zero
water 1evel and increases to 0.2% at 40 inches of water level and 0.7% at

G-16
at 57 inches. Carryover is also affected by the homogeneous velocity through the
nozzles. For example from Figure G-17, carryover increases from negligible amounts
at water levels below 20 inches about 20% at water levels greater than 40 inches.

Air/water tests were conducted on the radial vane separator. Even though there is
not a one to one correspondence between the size of entrained air bubbles and steam
bubbles, the trends should be the same. Figure G-18 shows that the amount of
unseparated air, leaving the vanes and thus carryunder, decreases with increasing
homogeneous nozzle velocity. The analysis in Reference 18 indicated similar re­
sults.

The radial vane separator is not recommended for further development for two rea­
sons . First, the prototype separator has not been proof tested at operating
conditions. (The scaled down radial vane separator has been tested.) Because of
the prototype separator's large size, a test facility with enough capacity for
proof testing is not readily available. Second, the allowable steam load appears
to be lower than other separators such as those recommended for further develop­
ment (Appendix F). In Reference 16, radial vane separators were sized for 300 MWt
and 1000 MWt Boiling Water Reactors. For the 300 MWt system, the allowable load
is marginal at about 32500 Lbm/hr. In Figure G-14 for the scaled down separator,
2
the permissible steam flow rate appears to be 45000 Lbm/hr or 10800 Lbm/hr ft .
2
The allowable steam load for the axial flow GE Separator is about 60000 Lbm/hr ft..

G-17
Downcomer

Upper Vortex
42

Lower Vortex

Base Plate

Figure 6-14. Radial Vane Separator.

G-18
1/8"

Vane Pair ; 135

Flow Passage

Nozzle

Figure G-15. Vane Pair.

G-19
1.4

TOTAL FLOW = 450,000 Lbm/hr 40


INLET QUALITY = 10%
CARRYUNDER

CARRYOVER
30

0.6
%

%
20

10

WATER LEVEL {inches above base plate at r = 9 inches)

Figure G-16.Typical Carryover and Carryunder Curves as a Function of Water Level


in FCRV-2.
50

40
CARRYOVER

30
%

20

10

10 20 30 40 50 60 70 80 90 100
HOMOGENEOUS NOZZLE VELOCITY (ft/sec)

Figure G-17. Carryover Before Break Point versus Homogeneous Nozzle Velocity.
Leaving Vanes

°\
o
V
\
\
\
A ir

\
-J S
Unseparated

//

s
0 X
"O o

°o ------ y~~
%
Volume

Homogeneous Nozzle Velocity, ft/sec

Figure G-18. Influence of Nozzle Velocity on Separator.

G-22
Section 7

TURBO SEPARATOR (Combustion Engineering)

The description and operation of the turbo separator (Figure G-19) is presented
in References 20 thru 22. The steam and water mixture rise through the support
castings around the hub through the spinner blades. The centrifugal force
exerted,on the steam and water mixture by the spinner blades cause the water to
be slung against the inside of the inner jacket wall and moisture laden steam
flows to the core. The moisture laden steam rises through the exit nozzle into
the secondary separator composed of corrugated plates. The corrugated plates
remove the moisture, which then flows down the outer jacket, and the steam rises.
The lip of the exit nozzle skims the water flow, and the top plate of the exit
nozzles redirects it into the downcomer. The water exits from the bottom of the
downcomer.

For a 7.375" exit nozzle diameter and a pressure of 1000 psi, the performance of
this separator is as follows:

1. Carryover is less than 0.05% up to steam flow rates of 25000 Lbm/hr


and a 9.3% inlet quality.

2. For steam flow rates between 25000 and 35000 Lbm/hr and an inlet
quality of 7.5%, carryover is in the .1 to .2% range.

3. For #1 and #2 conditions, carryunder is less than 0.5%.

4. At 5.3% primary outlet moisture (CO), the secondary separator be­


comes overloaded.

When the exit nozzle diameter is decreased below 7.375", the carryover perfor­
mance improved. Carryover below 0.25% for 10% inlet quality and water flow rates
of 369000 Lbm/hr was obtained up to steam flow rates of 41000 Lbm/hr.

This concept has the major advantage of compactness. It was not chosen because
its carryover performance for a given steam flow rate was lower than that of
several other separators such as the Design 5 Separator (Figure B-31). The outside
diameter of this separator limits placing more of these separators on the support
deck to achieve lower steam flow rate.

G-23
Steam

Secondary Separator

Moisture
Partial Section "A-A1
Jj A Moisture .
Hook ael Vaden i Water
ii \ Steam
liExtt Nozzle

— Downcomer

Outer Jacket^. Spinner Blades

Inner Jacket
i Hub

Water

Support Castines

Steam and Water Mixture

Figure G-19. Turbo Separator.

G-24
Section 8

DOWN-FLOW SEPARATOR (Babcock and Wilcox)

The down-flow separator (Figure G-20) is described in Reference 30. The steam
and water mixture is admitted through a tangential inlet and flows into an upper
chamber. Spaced from the upper chamber wall is a skimmer baffle which provides
an elongated nozzle at the tangential entrance of the separator, and the back edge
of this baffle acts as a skimmer to divide the water along the wall from the steam.
Because of the tangential entrance, whirl is imparted to the flow. Centrifugal
force separates the steam from the water as the mixture traces a helically down­
ward path due to gravity. The water concentrates along the wall and the steam
in the center. The circular end baffle prevents the steam from flowing downward
and allows the water to flow through the annular passageway into the lower chamber.
The impeller blades in the annular passageway covert the energy to a downward
energy which results in the water travel 1ing more quickly to the drum and allows
it to overcome the pressure caused by the high water levels within the drum. Be­
fore the water reaches the drum, it flows through collecting baffles which direct
the water radially inward. The vapor on the other hand passes upward through the
pressure distributor into the steam space. The collar at the top of the upper
chamber prevents water from overflowing into the steam space.

In Reference 30, a similar downflow separator is described. The major differences


are that the pressure distributor in Figure G-20 is omitted. Single or double
rows of these downflow separators are installed in boiler drums. Reference 31
considered carryover performance acceptable for the loading values given in the
following table.
Steam Load
Pressure (psi) Single Row (Lbm/hr ft) Double Row (Lbm/hr ft)

100 4000 600


1050 9000 15000
2500 15000 30000

G-25
The reference states that this downflow separator in combination with its dryer
system provides pure steam over a wide range of operating conditions. Reference 31
describes a separator similar to the one in Figure G-20 and Reference 30. This
separator according to the reference is from 10 to 12 inches in diameter. Using a
12 inch center spacing the length term can be eliminated from the table, and the
steam loading for the downflow separator in Reference 31 can probably be assumed
to be in the order of the values given in the table. Combustion Engineering tests
of downflow separators indicate that the allowable steam load to be low as indicated
in the table.

A downflow separator similar to the one in Figure G-20 and References 30 and 31
is being used successfully in boiler drums in many nations. The reason it is not
recommended for further development is that it has a low allowable steam load.
Other separators have higher allowable steam loads of from 45000 Lbm/hr to 75000
Lbm/hr and are more suitable for application to nuclear steam generators.
i
i
Pressure
Distributor
Collar
i!L _ ~

Water
Steam
Upper
Chamber

Skimmer
Baffle
Circular End
Baffle )

flange
Annular
Passageway Tangential Inlet

\ K /

Lower
Chamber

Collecting
Baffles
(spout)
UJ Pin

Figure G-20. Downflow Separator.

G-27
Section 9

BWR DOWNFLOW SEPARATORS (AEG and Allis Chalmers)

AEG and Allis Chalmers have tested BWR downflow separators. The operation and
description of those separators follows (References 23 through 29):

SEPARATOR OF FIGURE G-21A

The steam and water mixture rises from the reactor core and is redirected by
crossover cowls to two or three rows of separators located around the periphery
between the reactor pressure vessel and core shroud. The mixture flows down an
annulus through swirl vanes. By centrifugal action a water vortex occurs along
the periphery of the barrel with steam in the core of the barrel. The steam rises
and exits through a center pipe which directs it to a steam space below the dryers.
Because of gravity, the water flows downward through a perforated cone into the
reactor vessel downcomer to be recirculated. The perforated cone serves two func­
tions:

1. Stopping the 1iquid rotation

2. Closing the 1ower end of the vapor vortex in the central zone of
the separator.

SEPARATOR OF FIGURE G-21B

The steam and water is directed by an entrance cowl to a 45° tangential entrance.
The tangential entrance causes whirl and thus separation. The steam exits upward
through the center pipe and the water downward through the barrel.

SEPARATOR OF FIGURE G-22

The steam and water mixture enters and flows through the inlet pipe. Guide vanes
in the crossover cowl redirect the mixture to the annuli between the outtake
barrels and the steam offtake pipes. Swirl vanes at the bottom of these annuli
create the centrifugal force which separates the steam and water. The water flows
downward and exits through the four outtake barrels. Orifices at the bottom of
the outtake barrels offer appropriate resistance to the water flow. Steam rises
through the steam offtake pipes. Skimmers located at the top of these barrels
pick-off the moisture that collects along these pipes.

G-28
SEPARATORS OF FIGURE G-23

The steam and water mixture gains whirl as a result of their tangential entrance
through the scrol1 inlets. Water forms a vortex along the barrel wall and due to
gravity f1ows downward through the lower exit nozzle. Steam rises through the
upper exit nozzle.

Both air/water and steam/water tests at 1015 psia were conducted for the separator
in Figure G-21A. The carryunder performance curves shown in Figure G-24 indicates
that there is only a siight difference between the air/water and steam/water re­
sul ts. For carryover, the results do not agree as well. As shown in Figure G-24,
3
for a volumetric water flow rate of 6495 ft /hr and a volumetric steam flow rate
o 3
of 10590 ft°/hr, carryover is 265 ft /hr and for the same values of air/water
3
carryover is 160 ft /hr.

Performance curves for the separator in Figure G-22 are shown in Figure G-25.
Carryunder increases as the steam vortex in the outtake barrels increases. On
the other hand, carryover decreases for the same reason. Carryunder decreases
as water flow decreases. Both carryunder and carryover increase with increasing
water flow rate probably due to a finer and greater spray being generated at the
higher water flow rates.

AI1is Chalmers conducted air/water and steam/water tests in the development of


their downflow separators with sidescrolls. Figure G-26 shows carryunder air/
water test data for the separator developed for a BWR. Above 2800 GPM of water
carryover increases exponentially with water flow rate. Figure G-27 shows carry­
under and exhaust moisture performance for pressures of 600, 1200 and 1800 psig.
Carryunder increases as pressure increases and as water flow rate increases. There
appears to be a siight increase of exhaust moisture with increase of pressure.

AEG and Allis Chalmers in comprehensive test programs have developed downflow
separators for use in Boiling Water Reactors. An advantage of the downflow
separator is that they can be installed in a peripheral zone around the reactor
cone so that the space above the core would be free for easy access in fuel re­
placement operations. In the Lingen (KWU) Boiling Water Reactors, separators
similar to the one in Figure G-21B were used. In the Wurgassen (KWU) Boiling
Water Reactors, separators similar to the one in Figure G-21A were placed three
deep in the peripheral zone. It would be difficult to employ more than one row
of the separator shown in Figure G-21B in the peripheral zone because the centri­

G-29
fugation of this separator is developed by the tangential entrance and the design
would be prohibitive. Gardner reported in Reference 22 that separators in the
Wurgassen BWR experienced difficulties with the distribution (preferential flow
to a few separators causing them to overload) of the steam and water, and the
plant had to be downgraded to 65% full load. To increase the allowable flow load,
a downflow separator similar to the one in Figure G-22 was developed. This
separator was designed to be installed over the core. Allis Chalmers developed
a downflow separator to be installed in the BWR (Figure G-23). This separator
has side scrolls for tangential entrance of the steam and water. It is not
certain whether this separator was ever installed.

The downflow separator is not recommended for further development, since the
pressure drop of this type separator is usually high due to the steam flow
making two 180° turns in the in separation process. On an area basis, the
maximum allowable flow rate to this type separator was found to be about
20000 Lbm/hr ft^. Upflow separators have allowable steam loads up to 60000
Lbm/hr ft^. High velocities in the steam outlet pipe (center) would result in
high vapor velocities resulting in low allowable steam load.

G-30
Steam
Water

Center Pipe

Crossover
Steam * SCowl

Mixture Entrance Cowl


15.7

Center
Pipe
Steam
._Swi rl
Vanes Water
__________ _

Barrel Tangential Inlet


at 45°
31.5"

v Barrel
Core
Shroud

Perforated
11.8"

Cone

Figure G-21. AEG Cyclones.

G-31
I i
I I
I I

Skimmer

Crossover Cowl

Guide Vanes

Swirl Vanes
112.2"
Steam Offtake
Pipe

Outtake Barrel

Water Outflow
Orifice

Inlet Pipe

Figure G-22. AEG Separator with Four Outtakes.

G-32
A
Upper Exit Nozzle

Water
Steam

- - Scroll Inlet

Scrol1 Inlet

Barrel

Lower Exit Nozzle

G-33
( 1-FRACTION STEAM OR AIR CARRIED UNDER)
10943

6707 FT /HR
AIR OR
STEAM
13414 Fr7v

STEAM
AIR

WATER FLOW. FT /HR

3883 FT /HR 7801 FT /HR


WATER CARRYOVER. FT /HR

WATER

1130 FT /HR

STEAM
AIR

1 7060 10590 14120


STEAM OR AIR FLOW. FT3/HR
b.
Figure G-24. Performance Curves for Separator of Figure G-21A

G-34
0.3

CARRYUNDER
% OF WATER
0.2
X
• □
0.1
O -k§- JL J

0 8.8 17.6 26.5 35.3 44.1 52.9


STEAM FLOW,

a.
STEAM

Water Flow Lbm/hr x 10


% OF
CARRYOVER.

-f—a-------
17.6 26.5 3!
STEAM FLOW. LBM/HR

b.

Figure G-25. Performance Curves for Separator of Figure G-22


Conditions:
Water Level = 2.5' Submergence
Air flow -120 CFM" '

i
f
<D
E
rs /
o l
>
>> 1
i
S- 1___

Q) /
“D
C /
Z3
>5
o S-
/
GO O /
■/
/
/
**
y

_ - *

0 400 800 1200 1600 2000 2400 2800 3200


48140 1444200

Water flow = GPM Lb/hr

Figure G-26. Air water test of Allis Chalmers Separator.


Conditions:
System Pressure = 600 psi
Water Level = 12" to 30"
0 = Steam Exhaust
A = Carryunder

m
x
C arryunder-Percent Volume

398500 797000 1115800

Water Flow - GPM, Lbm/hr

Figure G- 27. Performance at 600 psi.


.Conditions:
System Pressure = 1200 psi
Water Level = 12" to 14"
0 = Steam Exhaust
A = Carryurider
8£ ” 9

o o Ct,c« O 0 o O 9d%o
A
A
A
I A A
& A
I A £ &A A
-------- A- ----A---- Z

359400 718800 1006320

Water Flow - GPM Lbm/hr

Figure G-28. Performance at 1200 psi,


Conditions:

System Pressure = 1800 psi


Hater Level = 12" to 30"
Steam Flow = 1500 to 3100 Lbm/hr
0 = Steam Exhaust
A = Carryunder

<
!° 00 Or* )
G"39

i i
o
o

0 i A
o

^ A AA
"SI
<

,A , A
V

A “a aa
A A

324500 649000 908600

Water Flow - GPM, Lbm/hr

Figure G-29. Performance at 1800 psi.


Section 10

AXIAL FLOW VENTURI SEPARATOR (Thesis by Prins)

(See Figure G-ll)

This concept was proposed by Prins (Reference 32). The steam and water mixture
enters the 4 inch inside diameter inlet section and flows through 8 flat guide
vanes with a 0° inlet angle and a 45° outlet angle. The guide vanes create
centrifugal force to separate the steam and water. The water forms a vortex on
the converging section of the vortex barrel and the steam migrates to the center
of the barrel. The converging section of the vortex barrel assists in the separa­
tion by increasing the rotation. The steam exits through the 1.3 inch inside
diameter vortex finder. The water flows by the 2.4 inch inside diameter throat
into the diverging section of the vortex barrel. Part of the pressure loss is
recovered in the diverging section of the vortex barrel by the kinetic energy being
converted into pressure head. The water exits through the water discharge passage
into the water pool.

The axial flow venturi separator was chosen as one of the promising concepts for
two reasons. First, this concept has a smaller outside diameter than most other
concepts evaluated. The smaller diameter results from this separator not having
an outer shroud. The diameter of the axial flow venturi model separator tested is
4 inches, and the prototype diameter recommended for use in 8 inches. Second, this
separator has the potential for low (less than 0.2%) carryunder because of its low
downcomer velocity.

The model 4 inch outside diameter axial flow venturi separator was tested in an air/
water loop. The carryover and carryunder results (Figure G-30) were plotted in
terms of a carryunder efficiency (rip), a carryover efficiency (n E), and an overall
efficiency (n q) .

n[) = 1 - 0D/eR

nE = 1 - (i-eE)/(i-eR)

\ = wcu/ws - (Ww - Wc0)/Ww

G-40
Where = Void fraction in downcomer

3^ = Void fraction in riser

= Void fraction above separator

Wcu = Carryunder mass flow rate

W$ = Steam mass flow rate

W = Water mass flow rate


w

WCo = Carryover mass flow rate

The abscissa of Figure G-30 and carryunder and carryover efficiencies are in terms
of volume qua!ities because there is a large difference between the ratio of air/
water densities compared with the ratio of high pressure steam/water densities.
Volume quality is more representative than mass quality. The curves on Figure G-30
without data points are for an allowable downcomer quality of 0.2% and exit quality
from the separator of 90%. The curves with data points are for measured values.

The axial flow venturi separator is not recommended for development because above
32% inlet volumetric quality, entrainment problems occur. No high pressure steam
and water test data was available. For most PWR steam generators, volumetric
quality is much higher.

G-41
100

nE ------------ A

nC---------- °
nD---------- V

6.6 ft/sec.
fic ie n

8.2 ft/sec.

9.8 ft/sec.

_J--------------------- 1______________ L------


10 20 30
% Void Fraction in Riser

Figure G-30. Efficiencies of Axial Flow Venturi Separator.

G-42
REFERENCES
1. G. C. Gardner. "Separation of Liquids from Gases and Vapors." Heat
Transfer and Fluid Flow Service. DR 46, AERE-R8917, 1977.

2. E. Blum, A. Si evens, and D. Wagner. "Steam Generator Design and Develop­


ment Capability." Curtis-Wriqht Corporation. 1977.

3. A. Si evens and E. Blum. "Performance Study." EPRI S123-1. 1978

4. P. F. Sokolowski. "Vertical Steam Drum." U.S. Patent 3751886. 1973.

5. R. R. Bennett and N. N. Kondic. "Momentum Flux Model for Liquid-Vapor


Separation." Proceedings of 17th National Heat Transfer Conference. 1977.

6. E. L. Burley. "Performance of Internal Steam Separation System in Boiling


Water Reactor". ASME Paper 69-WA/NE-24. 1969.

7. J. T. Cochran. "Axial Flow Vapor-Liquid Separator." U.S. Patent 3216182.


1965. “ ““

8. J. T. Cochran. "Pressure Recovery Axial Flow Vapor-Liquid Separator."


U.S. Patent 3329130. 1967.

9. C. H. Robbins and et al. "Gas-Liquid Separator." U.S. Patent 3603062.


1971. ' '

10. S. Wolf and R. H. Moen. "Advances in Steam-Water Separators for Boiling


Water Reactors." ASME Paper 73-WA/PWR-4. 1973.

11. R. H. Moen and et al. "Advances in Boiling Reactor Steam Separation


Systems." ASME Paper 69-WA/NE-S. 1969.

12. T. Stern. "Improvements in and Relating to Vapor-Liquid Separating


Apparatus." British Patent 926281.

13. A. A. Kudirka, C. H. Robbins, and R. H. Moen. "Gas-Liquid Separation on


a Vane of a Radial Vane Separator." ASME Paper 66-WA/NE-ll. 1966.

14. C. H. Robbins. "Analysis of the Radial Separator." GEAP-4011. 1962.

15. A. A. Kudirka. "Radial Vane Steam Separator Development: Experimental


Study of Carryunder Sources in Radial Vane Separator." EURAEC-1585.

16. R. H. Moen. "Nuclear Superheat Project Internal Steam Separation Develop­


ment of Radial Vane Separators." GEAP-4272. 1963.

17. R. H. Moen, A. A. Kudirka. J. E. Gifford. "Radial Vane Steam Separation


Development." GEAR-5283.

18. V. C. Hall and et al. "Air-Water Separation Test Program." CEND-168.

19. T. Ravese. "The Application and Development of the Turbo Steam Separator."
Combustion. 1954.
2Q, T. Ravese. "Vapor-frotn-Liquid Separation Apparatus." U.S. Patent 2648397.

21. H. D. Ongman. "Steam-Water Testing of Two Types of Combustion Engineering


Incorporated Separators." GEAP-3501.

22. G. C. Gardner. "Separation of Liquids from Gases or Vapors." RTFS DR 46.


1977.

23. K. H. Grabener. "Operating Conditions of the Downflow Cyclones for Steam/


Water Separation (In German)." Brennst-Warme-Kraft. Vol. 24, 1972.

24. U. Simon. "Theoretical and Experimental Determination of Behavior of


Cyclones in Steam Separation." Konstruction. Vol. 23, No. 11, 1971.

25. Tischer, VolIradt, and Kirchemajor. "Studies of Cyclones for Steam


Separation in Boiling Water Reactors." EURAEC-2127.

26. "Primary and Secondary Separation of Steam from Water by Mechanical Means."
ACNP-65003.

27. J. J. Ginoux (Z. Rouhani). "Two-Phase Flows and Heat Transfer." Hemisphere
Publishing Corporation. Chapter 13, 1978.

28. G. Poliak. "Construction of Steam Water Separators and Steam Dryers."


Konstruction. Vol. 23, No. 11, 1974.

29. Babcock and Wilcox Company. "Vapor-Liquid Separating ApparatusBritish


Patent 817650. 1956.

30. Babcock and Wilcox Company. "Vapor-Liquid Separating Apparatus." British


Patent 817650. 1956.

31. J. Karpisek. "Centrifugal Built-in-Water Separators for Heavy Duty Boiler


Drums." Strojirenstvi. Vol. 18, No. 9, 1968, pp. 649-655.

32. C. A. Prins. "Aspects of Two-Phase, Gas Liquid Separation Related to Nuclear


Steam Supply Systems." PhD Thesis at Technische Hogeschool Delft. 1971.

G-44
APPENDIX H

RECOMMENDED TEST PROGRAM


ILLUSTRATIONS

Figure Page
H-l Vanes H-18
H-2 Summary of Experimental Vane Design H-19
H-3 Converging Length Versus Exit Angle H-20
H-4 Lower Skimmer and Restrictor H-21
H-5 Calibration Curves H-22
H-6 Upper Skimmer and Straightening Vanes H-23
H-7 Screen Assembly H-24
H-8 Deflector H-25

H-i
Section 1

INTRODUCTION

This appendix outlines a test program to further develop each of the recommended
separator concepts described in Appendix F. If undertaken, this program would
extend the knowledge of steam/moisture separation devices for use in recirculating
steam generators.

In general, a three phase approach to testing is recommended to develop each of the


separator concepts. The first phase of testing would be a series of screening tests
for comparative evaluation of design features which might be included in that separa­
tor concept. In the second phase, further screening tests would be performed on
several separator geometries composed of the design features determined to be more
attractive during the first test phase. After selecting an optimized separator
design based on results of the second screening test phase, proof tests would be
performed at actual steam generator operating conditions in the third phase.
Particularly during the screening test phases, several retest cycles may be de­
sirable based on data developed as testing proceeds.

The moisture carryover, steam carryunder, and irrecoverable pressure loss per­
formance of design features and individual separator designs should be determined
for variations in vapor flow rate, liquid flow rate, and water level external to
the separator. Flow visualization is desirable, particularly during screening
test phases, to assist in determining if the design features operate as expected.

For the centrifugal vane, up-flow, down-flow separator, development should be


specifically directed toward achieving high flow loading per unit area, reduction
of the outside diameter for insertion through the steam generator manways, and
determining its carryover and carryunder sensitivity to flow and water level varia­
tions .

Development of the centrifugal vane, perforated jacket separator should take


advantage of its very low carryover at low flow loadings by testing the separator
in stacked arrangements as discussed in Appendices D and F. Development should also
explore substitution of slits or castellations for circulation perforations to im­
prove its performance at high flow loadings, as discussed in Appendix F.

H-l
Testing of the axial flow separator should attempt to enable a height reduction
from approximately 90 inches to 50 inches, and to allow an increased inlet quality
from its present limit of approximately 15% to 25% or greater.

Recommended general requirements for a test program are presented in this appendix.
For each type of separator recommended for further development, a description of
what testing is desirable is presented together with a brief explanation as to why
the testing is desirable. Further explanation of why additional development is
desirable may be found in Appendix F. The critical design features of each type
of separator are identified, and recommendations are provided concerning how those
features should be varied during testing. Nominal dimensions for most design
features are provided based on judgement gained during the course of this study.
These dimensions can be used only as a starting point for optimization. A vest
matrix listing specific variations in geometries and flow conditions has not been
included because the scope of study of any test program would be dependent on the
specific needs and objectives of the test, capabilities of the specific test
facility, and the funding available at the time of the tests.
Section 2

SCREEN TESTING CONSIDERATIONS

WORKING FLUID FOR SCREENING TESTS

Several fluids may be considered for screening tests. Each has its advantages
and disadvantages relative to accurately predicting actual steam separator per­
formance in the steam generator. Because the purpose of screen testing is to
evaluate the performance of a broad range of designs and features relatively
quickly and inexpensively, air/water at atmospheric pressure is the fluid re­
commended for the screening tests. Each fluid considered is discussed below.
Table H-l provides a comparison of characteristic parameters for each of the
fluids considered.

1. Steam/Water at 1000 psia

Steam/water at 1000 psia has the very strong advantage that the
separator concepts would be tested very close to operating condi-
tions. This fluid is not recommended because the costs of test
faci1ities and fuel for a large number of screening tests is pro­
hibitive.

2. Low Pressure Steam/Water (400 psia)

Low pressure steam/water has the advantages that the same fluids
would be used for the screen tests and proof tests. This is not
recommended, however, because there is a vast difference in the
properties of high pressure and low pressure steam/water mixtures,
as can be seen in Table H-l. Fuel and test facility costs would
also make a large number of screen tests extremely expensive.

3&4. Freon 12 and Freon 22

Gardener, et al,in Reference 3 performed separator tests using


Freon-12 to simulate steam/water at a pressure of 2600 psi. They
surmised that certain mechanisms in bubble dynamics were important
in the separation process. They indicated that bubble size was
dependent upon the dynamic force and the force of surface tension
acting on the bubble. In the centrifugal force field, the liquid
will tend to go radially outward and the bubble toward the center
of the separator. There could be a tendency, however, for the flow
to drag the bubble with it. Considering the dynamic force, force
of surface tension, centrifugal force and drag force, they indi­
cated that the following dimensionless parameters should be nearly
equal for model and prototype tests:

H-3
2
pf yf

Where = Liquid Density

= Liquid Viscosity

Ap = Density Difference Between the Phases

a = Surface Tension

As can be seen in Table H-l, both Freon-12 and Freon-22 at low


pressure have nearly equal C and CT values for steam/water at
2600 psia. Compared to steam/water at a pressure of 1070 psia,
however, the CT values for Freon-12 and Freon-22 differ by a factor
of 2.4 to 3.7, and the C values by 8 to 20. For air/water at
atmospheric pressure andvsteam/water at 1070 psia, the difference
in the Cy values differ by a factor of roughly 3.0, which is com­
parable to Freon-12 and Freon-22. A strong reason for not re­
commending Freon-12 or Freon-22 does not exist other than Freons
would be more costly, for example requiring a leak proof test sy­
stem.

5. Air/Water at 14.7 psia

Air/water is recommended as the test fluid because it is relatively


inexpensive and readily obtainable. Air/water has been used
successfully by many researchers in developing separators.
Following is a 1ist developed during this study of working fluids
previously used by various researchers for screening tests.

Researchers Working Fluid

Combustion Engineering Low pressure steam/water


General Electric (Reference 1) air/water
Kraftwerk Union (Reference 2) air/water
Gardner (Reference 3) Freon-12
Giho (Reference 4) air/water
Traube (Reference 5) air/water
Brandes (Reference 6) air/water
EURAEC (Reference 7) air/water
Dement'ev (Reference 8) air/water
Low pressure steam/water

H-4
COMPARISON OF SCREEN TESTS AND PROOF TESTS

During the screening tests, air/water should be used to identify undesirable fea­
tures . If a certain separator design does not operate adequately during the
screening tests using air/water, past test experience has indicated that the sepa­
rator design will probably not perform adequately at the high pressure steam/water
conditions. However, adequate performance of a separator design in air/water screen
tests does not ensure that the separator design will work in high pressure steam/
water conditions. For this reason, proof tests are necessary after a separator
design has been developed during air/water screening tests.

Volumetric flow rate instead of mass flow rate is recommended to define the test
points for the screening tests and proof tests. Volumetric flow rate is recommended
because the density of air is substantially different from the density of steam.
Screening test results with volumetric flow rate are more represent!ve and have a
closer one to one correspondence with high pressure steam/water results than mass
flow rate would have.

In choosing the points for a test matrix, it is recommended that the ratio between
the vapor- flow rate and 1iquid flow rate be duplicated for both the screening tests
and proof tests. The test matrix should be arranged in this manner in an effort
to maintain the same flow regime in the screening and proof tests.

H-5
Section 3

RECOMMENDED INSTRUMENTATION

In order to properly assess the total performance of a steam separation system, it


is recommended that instrumentation be provided to measure the following parameters.

1. Moisture Carryover

If a dryer is incorporated in the test, then moisture carryover be­


yond the dryer should be measured. This will give an indication
of the amount of moisture leaving the steam generator for that test
condition.

2. Exit Moisture from the Separator

This gives a direct indication of how well the separator is per­


forming. The inlet moisture to the dryer is a prerequisite in
determining the amount of dryer area needed.

3. Steam Carryunder

Steam bubbles entrained in the water leaving the separator decrease


the density of the recirculating fluid and hence reduces the recir-
culation ratio in the steam generator.

4. Total Separator Pressure Drop

This measurement is a direct indication of separator performance.


This information is needed in determining steam generator circula­
tion ratio.

5. Inlet Sudden Contraction Loss

This measurement will give the pressure loss from the steam genera­
tor riser to the separator standpipe.

6. Pressure Loss Along Standpipe

This measurement will give the pressure drop in the separator


standpipe. This measurement may also be used to calculate an inlet
quality which could be used as a check on the quality calculated
from the inlet flow rates.

7. Pressure Loss Across the Vanes

Most of the pressure loss in the separator occurs in the vanes.


This data may be used in optimizing the vane design.

H-6
8. Pressure Loss from Above the Vanes to the Exit Nozzle

This pressure drop may be used to estimate the height of the water
film inside the separator.

9. Pressure Loss Across the Exit Nozzle

If these pressure loss is too large carryunder may tend to be


excessive.

10. Water Level

The water level is a parameter which has been shown to affect the
separator's carryunder and carryover performance.

11. Steam/Water Interface

Knowing where the steam/water interface is will help to understand


the separator's operation.

12. Water Flow Rate

The water flow rate is a major parameter affecting separator per­


formance. Both the water flow rate into the separator and out of
the separator should be measured to enable a mass balance to be
calculated.

13. Steam Flow Rate

The steam flow rate is a major parameter in evaluating the separa­


tor' s performance. Both the steam flow rate into the separator
and out of the separator should be measured to enable calculation
of a mass balance.

14. System Pressure and Temperature

The system pressure and temperature should be measured to esta-


blish the fluid thermal physical properties.

15. Downcomer Flow Angle

The angle of the flow in the separator downcomer should be mea­


sured in order to design straightening vanes.

16. Visual Observation

The operation both inside and outside the separator should be


observed. This would be useful in better understanding the
operation of the separation and would be in judging how well the
separator or design feature performs. Visual observation is
particularly important during the screen test phases.

17. Bubble and Droplet Size and Distribution

This would be useful information in comparing the screen and proof


test results. This information would be particularly useful in
explaining the operation and performance of the separator, however
techniques for taking such measurements are somewhat developmental
and may be quite expensive.

H-7
Section 4

TESTING OF RECOMMENDED SEPARATORS

Centrifugal Vane, Upflow, Downflow Separator

Objectives in testing this separator concept should be to develop a compact


design tolerant of a wide range of water flow rates and water levels external
to the separator. The recommended procedure for optimizing this type of separa­
tor is to start with the basic Brandes concept (described in Appendix F) with
modified dimensions using the similitude analysis of Reference 6. The effect
of each dimensional characteristic on overal1 separator performance should be
determined during screening tests, and the optimum combination of dimensions
incorporated into a final centrifugal vane, upflow, downflow separator design
for proof testing. To perform adequately in a typical steam generator applica.-
tion, exit moisture from the separator should not exceed 5 percent, steam
carryunder should be less than 0.5 percent, and the allowable steam flow loading
for a separator less than 14 inches in diameter should be in the range of 75,000
Ibm/hr.

In the first screening test phase, air/water tests at atmospheric pressure should
be used to determine the effect of dimensional variations on separator perfor­
mance. To allow installation or replacement of the separator through a steam
generator manway, an outside diameter of 13.8 inches maximum should be an initial
constraint. Brandes tested a separator which had dimensions similar to the KWU
compact separator described in Appendix F. The similitude analysis of Reference
6 was used to scale down the Brandes separator, resulting in a recommended
nominal design for screen testing with critical dimensions as tabulated below.
See Figures F-4 and F-5).

¥ & Vane Spacing


df ds hs

Brandes Base
oo
h-4
oCXi
CO

Design 16.1 13.8 2.75 76°


o
o
CO

CXI
o

KWU Compact 15.5 12.6 2.4 10°

Recommended
Nominal Design 13.8 11.8 3.15 76° 25° 10°
During screening tests, the critical dimensions should be varied around the
nominal values to determine the sensitivity of separator performance indicators
to dimensional changes. The critical dimensions are 1 isted below, together
with the performance indicators to be observed or evaluated during the tests.

1. ds (Outside separator diameter, 13.8")

This dimension is smal1 enough to be passed the steam generator


manways, and should not be exceeded.

2. d^ (Outside vane diameter, 11.8")

The outside separator diameter with respect to the outside vane


diameter should be optimized in order that there be a smooth
transition of the water the inside of the outer jacket during the
separation process.

3. 3 (Curvature of Vanes, 25°)

Based on the outside vane diameter and number of vanes, there


should be optimum curvature for minimum pressure drop and maximum
separation.

4. ip (Exit vane angle, 76°)

The vane angle should be chosen Tor optimal separation.

5. hs (Vane height, 3.15")

There is probably a vane height which, if exceeded, does not im­


prove separation.

6. hfos (Distance from top of vanes to skimmer, 21.6")

This height should be great enough to allow a water vortex on the


inside of the outer jacket to form so that the water can be
readily skimmed.

7. hfus (Distance from bottom of vanes to downcomer exit, 9.2")

This height should be great enough to allow the separated steam


to rise without being carried under.

8. Wf (Distance from the base of the separator to the bottom of the


SKimmer, 32.7")

This dimension should be great enough to allow a low exiting water


velocity so that carried under steam bubbles can escape.

Evaluation of results from the first phase may indicate that additional
testing of dimensional characteristic variations are needed. The test
matrix chosen may not definitely establish break points in performance

H-9
curves, etc. Any additional testing should be completed before proceeding to
the second phase.

The second phase recommended in development of the centrifugal vane, upflow,


downflow separator would consist of air/water screen tests of candidate optimized
separator designs. After evaluation of results from dimensional characteristic
testing in the first phase, one or more separator designs having promising com­
binations of dimensional characteristics would be selected. These separator
designs would be tested with air/water to determine overall separator perfor­
mance. Modification to the separator design(s) should be made as indicated by
evaluation of interim results of this phase of testing. The second phase objec­
tive should be to select an optimized design based on air/water testing.

The third test phase should consist of proof tests of the optimized separator
with steam and water at steam generator operating conditions. Test conditions
should be selected to cover the ful1 range of pressure, water and steam flow
rates, and external water levels expected for the intended steam generator
application(s).

2. Centrifugal Vane, Perforated Jacket Separator

In its present configuration the centrifugal vane, perforated jacket separator


(Figure F-7) is a very compact separator with highly efficient steam carryover
performance. The basic objective in further development would be to produce
a separator design and arrangement which would result in almost negligible
moisture carryover. A secondary objective would be to reduce steam carryunder
at low external water levels.

Two approaches are recommended for further improvement of carryover performance.


The first would consist of testing the performance of separators stacked for
parallel flow paths as discussed in Appendix D. The second approach would
evaluate improvements in performance resulting in substitution of slots or
castellations for the present system of circular perforations, as discussed
in Appendix F.

The two approaches to improvement are recommended to be conducted in sequential


air/water screen testing phases. The first phase would evaluate the improvements
gained by the substitution of slots or castellations for circular perforations.
After the most nearly optimum design has been developed for this consideration,
it would be tested in the stacked arrangement.

H-10
In the first screen testing phase, various arrangements of perforations in
the separator barrel should be tested with air/water at atmospheric pressure.
The castellation modification and the slots modification concepts are shown
in Figures F-10a and F-lQb respectively. Numerous sizes and arrangements of
such modifications could be tested. It is recommended that at least three
combinations of size and arrangement for each type of modification be
initially tested. Observations and data evaluation from the initial tests
may indicate trends which would lead to additional configurations for
testing, until the most nearly optimum design is reached. The primary indi­
cator of performance would be moisture carryover. Steam (air) carryunder
would be a secondary indicator, particularly with respect to size and angle
of the slot modification.

If the first phase of screen testing does not provide positive indication
of significant performance improvements, the slots and castellation modifi­
cation is not recommended for further development in the second phase of screen
tests, or in the proof tests.

The second phase of screen testing recommended for the centrifugal vane,
perforated jacket separator would be air/water tests of the stacked, parallel
flow arrangement. The stacked arrangement inside a steam generator is shown
in plan view in Figure D-l and in elevation view in Figure F-9. For air/
water testing, a repeating pattern would be selected which would provide at
least one separator at each elevation which experienced typical flow patterns.
Other than indications of net performance of the stacked arrangement, two
considerations should be carefully addressed during these tests:
1. To reduce the steam drum height required in the steam generator
for the stacked arrangement, the recommended length of the separa­
tor has been arbitrarily reduced from approximately 50 inches to
30 inches. The tests should verify that the reduced length is
sufficient to fully develop centrifugal separation within the sepa­
rator for the range of flow rates under consideration. (Length
required for separation is directly proportional to flow rate,
which would be reduced in the stacked arrangement.) Adjustments
in separator length should be made if there are indications of in­
complete separation.

2. Since the discharge channel for the upper separator is quite long
before reaching the separator deck, there is a possibility of higher
steam carry under from the upper stack. Additional air/water tests
with adjustments in downcomer resistance may be required if excessive
carryunder from the upper stack is indicated.

H-ll
Successful indications from either or both screen testing phases should be
followed by a third phase of proof testing with steam and water at steam genera­
tor operating conditions. Testing of the slots or castellations modifications
alone can easily be accomplished in the conventional manner, testing a ful 1
scale separator over the range of pressure, steam and water flow rates, and
external water levels expected for the intended application. For most test
facilities, however, testing of the stacked arrangement at typical steam genera­
tor conditions may exceed the capabi1ities of the facility for steam or water
flow capacity. Two courses of action are suggested if the capacity of the test
facility is inadequate. Scaled down separators can be tested, using appropriate
sealing and similitude analysis. As an alternative, full-scale separators can
be tested in a simulation of the stacked arrangement, using steam flow through
only one separator at a time and providing water flow from adjacent separators
based on observations made during air/water tests.

3. Axial Flow Separator

The axial flow separators developed by General Electric operate with an inlet
steam quality of approximately 15 percent which is typical of BWR applications.
In PWR steam generators, inlet quality to the steam separators is in the range
of 25 percent. Consequently, one of the primary objectives in further develop­
ment of the axial flow steam separator is to increase the allowable inlet qua-
1ity to 25 percent or greater. In its present configuration (Figure B-10) the
General Electric separator is approximately 90 inches in overal1 length. To
make the separator suitable for backfit to typical PWR steam generators, another
objective in further development is to reduce the overall length to the vicinity
of 60 inches.

Figure F-12 shows a composite design concept based on the various axial flow
separators that have been used/or tested. The recommended approach to further
development is to perform air/water screen tests of separators with variations
in the design features shown in Figure 12. After evaluation of the performance
characteristies of variations in the design features, one or more axial flow
separator designs having promising combinations of design feature dimensions
would be screen tested with air/water at typical steam generator conditions
using the design(s) which appears to be optimum based on results of the first
two phases of air/water testing.

H-12
Following is a 1isting of the design features shown on Figure F-12 which should
be evaluated during the initial screen testing phase. A recommended nominal
value for dimensions which should be varied in the test matrix is given, along
with the indicators which should be observed or evaluated during the tests.

A. Vanes

The following dimensions should be considered in the design of the vanes.


Figure H-l shows these dimensions and 1ists nominal values for these
dimensions.

The following factors should be used in evaluating the performance of the


axial vanes. The first factor that should be considered is the thickness
and height of the water vortex on the inside of the jacket. The second
factor is an observation of the separated water and steam. Perhaps the
separation is not complete, or perhaps the water vortex has excess steam
bubbles, or the steam cone may have excess water droplets. The third
factor is the measured carryunder and carryover.

1. Iniet Transition

In order to maintain a nearly constant velocity from where the mixture


enters at the support plate to the f1ow area between vanes, the separa­
tor is designed with a taper at the inlet. This should minimize
pressure drop. Figure H-l shows the inlet transition.

2. Separator Diameter

A nominal separator diameter of 12" was chosen. Other diameters such


as 8" and 10" are recommended to be tested. The 12" nominal design
allows the separators to be passed through the nuclear steam generator
manways for a retrofit operation. Larger diameter separators would
also require larger mass flow rates and for this reason may be more
difficult to test in an existing facility.

3. Hub

The hub should be designed to be round at the inlet to minimize


pressure losses.

H-13
A review of axial vanes (Figure H-2) indicated that the hub to separator
diameter ratio was usually in the range of .25 to .35. If the hub to
separator diameter ratio was significantly larger then the pressure
loss would increase substantially. A reduction in hub to separator
diameter ratio would cause an increase in the converging length. The
thickness of the water vortex on the inner jacket could be great
enough that water might flow out the end of the separator.

4. Vane Height

A nominal vane height of 6" is recommended. Above a certain height


does not improve separation.

5. Inlet Angle

A 90° inlet angle is recommended to allow an adequate transition to a


30° to 50° deflection at the outlet angle (Appendix E).

6. Outlet Angle

A review of axial vanes (Figure H-2) shows that the outlet angle at
the hub varies from 50° to 67° and the outlet angle at the wall varies
from 30° to 40°. The commonly used outlet angle of 35° is recommended
for the initial test.

If the angle is too smal1, large centrifugal forces are produced re­
sulting in reduced bubbled size and therefore inefficient separation.
A corresponding large operation pressure drop is also produced. On
the other hand, a large angle results in insufficient centrifugal
forces and therefore inefficient separation.

As shown on Figure H-3 (Reference 8), the vane exit flow angle affects
the converging length. Converging length decreases as vane exit angle
decreases.

7. Number of Vanes

Six vanes are a compromise for initial experimentation. A large number


of vanes create an excessive pressure drop and too few vanes produce
inefficient separation.

H-14
B. Restrictor

Figure H-4 shows the restrictor arrangement. For minimun carryunder.


Figure H-5a indicates that the restriction clearance area should be between
50 to 60 percent of the pick-off ring area (Reference 10). The design
on Figure F-12 has a ratio of 58% which is recommended for initial experi­
mentation.

The restrictor throttles the water flow resulting in a water vortex thick­
ness greater than the lower skimmer clearance. It also results in the
separator downcomer being full of water. Optimizing the restrictor clear­
ance results in minimum carryunder. If the restrictor clearance becomes
too smal1, excessive carryover may occur. The factors that should be used
to evaluate the restrictor clearance are carryover and carryunder. The
operation should also be observed.

C. Lower Skimmer

Figure H-4 shows the lower skimmer. The first step in choosing the lower
skimmer dimensions is to optimize the ski niner (pick-off ring) clearance
with respect to the restrictor clearance to obtain minimum carryunder
(Figure H-5a). The second step is to optimize the end gap and overlap
length (shown on Figure H-4) for minimum carryover (Figures H-5a and H-5c).
The following nominal dimensions were chosen: skimmer clearance of 1",
end gap of 1", and overlap length of 0.5".

The lower skimmer channels the separated water from the separator to the
support deck. The lower separated water from the separator to the support
deck. The lower skimmer should be optimized to prevent steam from being
picked-off, and it should be optimized to handle a wide range of flow condi­
tions. The operation of the lower skimmer should be observed and carryunder
measured.

D. Upper Skimmer

The upper skimmer is shown in Figure H-6. This skimmer gap should be
optimized by flow testing. A nominal gap of 0.5" is recommended for initial
testing. The water that spills over the lower skimmer should be picked-off
by the upper skimmer. This skimmer should pick-off all the water with a
minimum of steam. The operation should be observed and carryover and carry­
under monitored. The objective of these tests is to calibrate the upper
skimmer to minimize carryover while not detrimentally effecting carryunder.

H-15
E. Straightening Vanes

The straightening vanes are shown in Figure H-6. Cochran, in Reference* 9,


designed straightening vanes with 5-15° entrance angle. The Kraftwerk
Union Design 4 separator (Appendix B) appears to have a 45-55° entrance
angle for their straightening vanes. It is recommended that the entrance
angle for the straightening vanes be optimized during screen testing.
Twelve vanes with entrance angles varying from 15° to 55° are recommended
for the screen tests. These vanes arrest the rotary motion of the water
and reduce the velocity at the discharge. The vanes may help to minimize
carryunder. Optimally, the straightening vanes should be designed so that
their entrance angle is the same as the directi on of the water flow swirling
in the downcomer.

F. Secondary Separators

Combustion Engineering, Inc., has tested two secondary separators, a screen


assembly (Figure H-7) and a deflector (Figure H-8). The screen assembly
had significantly better performance and is recommended for initial tests.

The purpose of the secondary separator is to extract the larger droplets


from the exiting steam flow and distribute the steam flow over the area
above the separator to prevent the steam from jetting into the dryers.
The secondary separator should be attached to the outlet of the separator.
To as much extent as possible, the operation of the secondary separator
should be observed. Other secondary separators which may be considered
for testing are discussed in Appendix E.

6. Secondary Separation Devices (Figure E-l)

Secondary separation devices may be installed in the downcomer of the


separator to separate the carried under steam from the exiting water. The
secondary separation devices should be installed in the annular separator
downcomer. If possible, the operation should be observed. The secondary
separation devices should be compared to determine which reduces carryunder
the most without having detrimental effects on carryover. The following
secondary separation devices are recommended for testing.

H-16
Secondary Separation Device Test Candidates

Test Device

1 Baseline test, no device

2 Baffle arrangement (Figure E-15)

3 Bubble rakes (Figure E-17)

4 Axially disposed vanes (Figure E-18)

5 Flow channels and vanes (Figure E-19)

6 Hollow vanes (Figure E-16)


Annular clearance would have to be
increased to allow for the hollow
vanes.

Because of the number of different design features to be evaluated during the first
screen testing phase, the preferred procedure would be to optimize the more domi­
nating features (vane design, restrictor clearance, and skimmer design) in sequence
before proceeding to the features anticipated to be of less influence on overal1
performance.

After evaluation of results from screen testing of design features, one or more
axial flow separator designs with promising combinations of design feature dimen­
sions should be selected. These designs would be subjected to a second phase of
air/water screen testing to evaluate overall separator performance. Modification
of these separator designs may be indicated by interim results during the second
screen testing phase. At the conclusion of this test phase, an optimum axial flow
separator design should be indicated.

The final test phase should consist of proof tests of the optimized design with
steam and water at steam generator operating conditions. Test conditions should be
selected to cover the full range of parameters (steam pressure, steam and water
fl ow rates, and water level external to the separator) expected for the intended
steam generator applications.

H-17
Inlet Transition

Nominal Vane Dimensions

Number of Vanes = 6
o

Inlet Angle
Or
o
II

Outlet Angle
O
LO

Inner
o
II

Outer
CO
o
H

Overlap = 15°

Hub = 30% Area of Cylinder

Figure H-l. Vanes.

H-18
Number of Inlet Angle Outlet of Hub/Diameter
Vanes of Vanes Vanes Ratio

Hub Wall Diameter Ratio

Allis Chalmers (Reference 12) 6 90° Unknown 40° 2.25710.25"

Traube (Reference 13) 4 900 500 30° 1.073.7"

Rouhani (Reference 14) 12 85o-90° Unknown 20o-40O .25"-.35"


H-19

Dement'ev (Reference 15) 6 Appears 90° Unknown 35° 1.173.9"

Dement'ev (Reference 8) 6 Appears 90° 67° 35° 1.173.9"

Cochran (Reference 11) 8 900 Unknown 300-40° Unknown

Giho (Reference 4) 12 Appears 90° Unknown 1.874.7"

Figure H-2. Summary of Experimental Vane Design.


Mean Tangential Velocity
a' =
Mean Axial Velocity

Converging
Length
(inches)

4
H-20

Band of Experimental Uncertainty

I _______________ i _ _______________ I
0.5 ... 1.0 175“
56.3° 45° 26.6°

VANE EXIT FLOW ANGLE, a

Figure H-3. Converging Length Versus Exit Angle.


Lower Skimmer
End Gap 1.0"
Overlap Length
T
0.5"

Restrictor

Swirler Assembly

Figure H-4. Lower Skimmer and Restrictor.

H-21
PERCENT OF DESIGN
FLOW RATE
O'

oc
§
l_U i
Q
g
§
£ 2:
o
2:

0.4 0.5 0.6 0.7 0.8 0.9

AREA OF RESTRICTION
AREA OF PICK-OFF RING

STEAM SEPARATOR TESTS


t 20 AT 1000 PSI and 546°F
£5 10 w/o INLET QUALITY
§ 18 STEAM SEPARATOR TESTS
^ 16
LlJ CQ AT 1000 RSI and 546°F
QC
ZD cfs 14 MIXTURE FEED RATE
J-
1000 CU. FT./HR
2 12 35
10 MIXTURE FEED RATE
1000 CU. FT./HR.
"
UJ
8 (100%)
1AINED MOISTURI

6 (91.5%)

4 (80%)
4 (69%)
2 20 (57%)

,6 -7 -S .7 /-O /./ 7.2 /. J


£ VORTEX TUBE END GAP-INCHES
LJ
fSa too 75 so 15
INNER TUBE (86) 0VERLAP-% OF END GAP

b. 1.0 1.2 1.4 1.6 1.8 2.0

LIQUID DISCHARGE AREA


INLET NOZZLE AREA
c.

Figure H-5. Calibration Curves.

H-22
Straightening Vanes

Figure H-6. Upper Skimmer and Straightening Vanes.

H-23
0.44"

4" to 6

0.313" dia.

Wire mesh, #6 x .045" dia. wire

Figure H- 7.

H-24
Deflector

Drain Pipes
Six-V' sch 40 pipes
H-25

Separator
Exit Nozzle!
Water

Steam

Figure H- 8, Deflector.
REFERENCES

1. C. H. Robbins. "Steam Separation Reduces BWR Size." Electric World, 1963


pp. 78,79.

2. R. Bouecke and G. Schucktanz. "Two Phase Flow Separation in U-Tube Steam


Generators." In Proceedings of 17th National Heat Transfer Conference, 1977.

3. G. C. Gardner, I. G. Crow, and P. H. Neller, "Carryunder Performance of Drums


in High Pressure Circulation Boilers." Proc. Instr, Mech. Engrs., Vol. 187.

4. H. Giho, "Experimental Studies of Steam/Water Separator.", NSJ-TR-62', 1966.

5. K. Traube and J. Vollradt. "Investigations on the Separation of Flowing Gas/


Liquid Mixtures by Centrifugal Force." Chem.-Ingr.-Tech., Vol. 34, 1962,
pp. 827-833.

6. H. Brandes. "Experimental Investigation of Centrifugal Separators - 1."


Energie and Technik, Vol. 23, 1971, pp. 283-290.

7. Tischer. "Studies on Cyclones for Steam Separation in Boiling-Water Reactors."


EURAEC 2127, 1968.

8. B. A. Dement'ev and et al. "Investigation of a Separator Unit with Axial


Supply of Gas/Liquid Mixture." Teploenergetika, No. 3, Vol. 19, 1972,
pp. 25-29.

9. Cochran. "Pressure Recover Axial Flow Vapor-Liquid Separator." U.S. Patent


3329130, 19.

10. R. H. Moen and S. Wolf. "Gas-Liquid Separator." U.S, Patent 3902876, 1975.

11. Cochran, Ongman, and Weber. "Axial Flow Vapor-Liquid Separator." U.S.
Patent 3216182, 19.

12. "Moisture De-Entrainment Tests in Two and Four-Inch Diameter Test Sections."
ACNP-5921, 1959.

13. K. Traube and J. Vollradt. "Investigations on the Separation of Flowing Gas/


Liquid Mixtures." Chem.-Ingr.-Tech., Vol, 34. 1962.

14. S. Z. Rouhani. "Separator for Mixture of Steam and Water." Swedish Patent
7309949-1.

15. B. A. Dement1ev and et al. "Experimental Investigation of Centrifugal Steam/


Water Separator." Teploenergetika, Vol. 19, No. 3, 1972.
APPENDIX I

SPACE/SIZE LIMITATION EVALUATION


ILLUSTRATIONS

Figure Page

1-1 Steam Generator No. 1 (Combustion Engineering, Inc.) 1-3


1-2 Top View of Design 2 Separator in Steam Generator 1-4
No. 1 (232" I.D.) (Westinghouse Separator in
Combustion Engineering Steam Generator)
1-3 Top View of Design 3 Separator in Steam Generator 1-5
No. 1 (General Electric Separator in Combustion
Engineering Steam Generator)
1-4 Top View of Design 4 Separator in Steam Generator 1-6
No. 1 (Kraftwerk Union Separator in Combustion
Engineering Steam Generator)
1-5 Design 2Steam Generator (Westinghouse Steam Generator) 1-7
1-6 Top View of Design 3 Separator in Steam Generator 1-8
No. 2 (159" I.D.) (General Electric Separator in
Westinghouse Steam Generator)
1-7 Top View of Design 4 Separator in Steam Generator 1-9
No. 2 (Kraftwerk Union Separator in Westinghouse
Steam Generator)
1-8 Top View of Design 5 Separator in Steam Generator No. 2 1-10
(Combustion Engineering Separator in Westinghouse
Steam Generator)
1-9 Steam Generator No. 3 (Kraftwerk Union Steam Generator) 1-11
1-10 Top View of Design 2 Separator in Steam Generator No. 3 1-12
(180" I.D.) (Westinghouse Separator in Kraftwerk Union
Steam Generator)
1-11 Top View of Design 3 Separator in Steam Generator No. 3 1-13
(General Electric Separator in Kraftwerk Union Steam
Generator)
1-12 Top View of Design 5 Separator in Steam Generator No. 3 1-14
(Combustion Engineering Separator in Kraftwerk Union
Steam Generator)

I-i
APPENDIX I

SPACE/SIZE LIMITATION EVALUATION

Four presently used separators were evaluated on a space/size limitation and


performance basis in three steam generators. In each case, the evaluation
considered that the original dryer system would remain in the respective steam
generator. The results are summarized in Table 1-1. The separator design type
numbers conform to those in Appendix B. For a different type separator to be
retrofitted, either the support mechanism of the separator to the standpipe and/
or the support plate would have to be modified. The separator designated as
Design No. 1 in Appendix B was not evaluated because the only available perfor­
mance data (345 psia) was much lower than the operating pressure of the three
steam generators (800-1100 psia). All of the separators in combination with their
respective dryer systems have carryover less than .25%.

The separators proposed for further development (Appendix F) fit through the
manways and into the available space for all three steam generators. Separators
should be developed for the specific application whether it be low or high operating
pressure or for certain loadings and qualities. The references for the three
steam generators and the five separator designs are given as follows:

Steam Generator No. 1 (A Combustion Engineering Design) Reference 3

Steam Generator No. 2 (A Westinghouse Design) Reference 1

Steam Generator No. 3 (A Kraftwerk Union Design) Reference 4

Design 1 Separator (A Curtis Wright Design) Reference 2

Design 2 Separator (A Westinghouse Design) Reference 5

Design 3 Separator (A General Electric Design) Reference 6,


Reference 7

Design 4 Separator (A Kraftwerk Union Design) Reference 4

Design 5 Separator (A Combustion Design) Reference 8

1-1
DESIGN 2 SEPARATOR

The Design 2 Separator was originally designed to operate in Steam Generator No. 2.
The performance of Design 2 Separator in other steam generators was based on an
analytical technique described in Reference 1; whereas, the performances of Designs
3, 4, and 5 separators were based on actual test data. The predicted exit moisture
from the Design 2 Separator was too high for both Steam Generators No. 1 and No. 3.
Figure 1-2 is a layout of Design 2 Separator in Steam Generator No. 3 (Figure 1-9).
As indicated in Table 1-1, this separator cannot fit through the manways of any of
the steam generators.

DESIGN 3 SEPARATOR

This separator was not originally developed for PWR steam generator application.
Design 3 separator has a very high allowable loading (flow rate) on an area basis,
but it is designed for an inlet quality of 10 to 15%. To be effective in any of
the steam generators studied, it would have to be redesigned for about 25% inlet
quality. This could possible be accomplished by modifying the restriction in the
separator downcomer. For Steam Generator No. 1, the height of this separator would
have to be reduced. Figure 1-3 shows the layout of Separator No. 3 in Steam Genera­
tor 1. Figure 1-6 and 1-11 show the layout of Separator No. 3 in Steam Generator
No. 2 and 3 respectively.

DESIGN 4 SEPARATOR

This separator was originally designed for use in Steam Generator No. 3. Design 4
Separator could not be retrofitted into existing Design 1 and 2 steam generators
because its diameter is larger than the manway opening. Increases in manway dia­
meters to at least 20 inches would be required. If Design 4 Separator could be
installed into Steam Generator No. 2 as shown in Figure 1-7, the carryover perfor­
mance should be less than 0.25%. For Design 4 Separators to be installed into
Steam Generator No. 1 as shown in Figure 1-4, either the separator deck would have
to be 1 owered or the dryer deck raised.

DESIGN 5 SEPARATOR

This separator was originally designed for use in Steam Generator No. 1. Design 5
Separator is smal1 enough to be passed through the manways of all three steam
generators, and it is short enough not to cause any height problems. Enough separa­
tors could be installed into any of the three steam generators evaluated to have
adequate performance. Figures 1-8 and 1-12 show the layouts of Steam Separator
No. 5 in Steam Generators No. 2 and 3 respectively.

1-2
Figure 1-1. Steam Generator No. 1 (Combustion Engineering, Inc.)

1-3
Figure I- 2. Top View of Design 2 Separator in Steam Generator
No. 1 (232" I.D.) (Westinghouse Separator in
Combustion Engineering Steam Generator)

1-4
Figure 1-3. Top View of Design 3 Separator in Steam Generator
No. 1 (General Electric Separator in Combustion
Engineering Steam Generator)

1-5
Figure 1-4. Top View of Design 4 Separator in Steam
Generator No. 1 (Kraftwerk Union Separator
in Combustion Engineering Steam Generator)

1-6
Figure 1-5. Design 2 Steam Generator (Westinghouse Steam
Generator)

1-7
Figure 1-6. Top View of Design 3 Separator in Steam Generator
No. 2 (159" I.D.) (General Electric Separator in
Westinghouse Steam Generator)

1-8
Figure 1-7. Top View of Design 4 Separator in Steam Generator
No. 2 (Kraftwerk Union Separator in Westinghouse
Steam Generator)

1-9
Figure 1-8. Top View of Design 5 Separator in Steam Generator
No. 2 (Combustion Engineering Separator in Westinghouse
Steam Generator)

1-10
4690

Figure 1-9. Steam Generator No. 3 (Kraftwerk Union Steam


Generator)

1-11
Figure 1-10. Top View of Design 2 Separator in Steam Generator
No. 3 (180" I.D.) (Westinghouse Separator in
Kraftwerk Union Steam Generator)

1-12
Figure I-11. Top View of Design 3 Separator in Steam Generator No. 3
(General Electric Separator in Kraftwerk Union Steam
Generator)

1-13
Figure 1-12. Top View of Design 5 Separator in Steam Generator No.
(Combustion Engineering Separator in Kraftwerk Union
Steam Generator)

1-14
REFERENCES

1. R. R. Bennett and N. N. Kondic. "Momentum Flux Model for Liquid-Vapor


Separation." 17th National Heat Transfer Conference. 1977.

2. D. Steininger. "EPRI Correspondence to W. R. Carson." 1978.


TM
3. H. K. Williams. "System 80 NSSS Steam Generator Design." American
Power Conference. 1973.

4. R. Bouecke and G. Schucktanz. Two Phase Flow Separation in U-Tube Steam


Generators." 17th National Heat Transfer Conference. 1977.

5. P, F. Sokolowski. "Vertical Steam Drum." U.S. Patent No. 3751886.


1973.

6. R. Moen. "Gas-Liquid Vortex Separator." U.S. Patent No. 3902876. 1975.

7. S. Wolf and R. H. Moen. "Advances in Steam-Water Separators in Boiling


Water Reactors." ASME Paper 73-WA/PWR-4. 1973.

8. R. Patterson, R. Micheller, and J. Carney. U.S. Patent 3614863. 1968.

1-15
APPENDIX J

HYDRAULIC EVALUATION
ILLUSTRATIONS

Figure Page

J-1 "CRIBE" Results Design B Steam Generator J-3


J-2 "CRIBE" Results Design B Steam Generator J-4
J-3 Shock Loss Factor vs. CR for Economizer Steam Generator J-5
J-4 Carryunder vs. CR Non-Economizer Steam Generator J-6
J-5 Carryunder for Economizer Steam Generator J-7

J-i
APPENDIX J

HYDRAULIC EVALUATION

The Combustion Engineering "CRIBE" code was used for the hydraulic evaluation.
"CRIBE" calculates steam generator pressure fields, mass inventories, and cir­
culation ratios. It is an extension to two flow paths of the single path "CRIB"
program developed by Bettis (Reference 1).

The "CRIBE" program was run for three steam generator models: Two non-economizer
units designated as A and B and one economizer unit. The units are typical of
many nuclear steam generator designs presently in operation. For both non-econo­
mizer units, only a moderate gain could be achieved by reducing the separator shock
loss factor. For non-economizer Steam Generator B with a separator shock loss fac­
tor of 9, Figure J-1 shows a circulation ratio of 4.1. With a reduction of shock
loss factor to 4.5, the circulation ratio increases by 0.27 to 4.37. Figure J-2
for non-economizer Steam Generator A illustrates similar results. For a shock loss
factor of 3.0, the circulation ratio is 3.9, and for a shock loss factor of 1.5,
the circulation ratio is 4.2. The "CRIBE" run for varying the separator shock
loss factor of the economizer steam generator was similar to those of the non­
economizer units. From Figure J-3 for circulation ratios of 3.3 and 3.6, the
shock loss factors are 3.0 and 1.5, respectively. Because only 20% of the pressure
drop in the circulation flow loop is from the separators, the gain in circulation
flow rate by reducing the separator shock loss factor is 1imited.

The shock loss factors for five of the presently used separators (Appendix B) were
compared and are shown in Table J-1. The following representative areas were
chosen so that the separator shock loss factors could be compared on a common basis.

1. Hexagonal Area - Separators are installed on a triangular pitch


giving each separator a equal hexagonal area to serve.

2. Total Area - A second area is based on the rise area (area of


support covered with separators) divided by the number of separators.

Comparisons using both areas in Table J-1 yield similar shock loss results. The
Westinghouse Design 2 Separator and the General Electric Design 3 Separator have

J-1
the lowest shock loss factors. This may be attributed to the use of axial vanes
in these designs.

Moderate gains in steam generator circulation ratio may be achieved by modifying


the vanes and entrance to the separator. An evaluation showed that in separator
designs 1, 4 and 5 with the centrifugal vanes, the flow makes a 90° turn before
being separated, which induces a relatively large pressure loss. Axial vanes
such as Designs 2 and 3 may reduce the separator pressure drop by a factor of
2 or more.

A small reduction in separator pressure drop may also be achieved by modifying


the entrance at the separator standpipe. The sudden contraction pressure loss
from the riser to the separator standpipe is on the order of 0.3 to 0.5 psia,
and could be substantially reduced by rounding the entrance of the separator
standpipe.

The hydraulic evaluation using "CRIBE" showed that carryunder has negligible
effect on the circulation ratio of a non-economizer steam generator and has
moderate effect on the circulation ratio of an economizer steam generator.
Figures J-4 shows that for 100% load, circulation is almost a constant 4.1
for 0 to 6% carryunder. At at 50% load, circulation ratio decreases from
8.45 for no carryunder to 8.00 at 6% carryunder. For an economizer steam
generator at full load. Figure J-5 illustrates the influence of carryunder.
At no carryunder, circulation ratio is 3.2, and for 6% carryunder, CR is 2.27.
A similar decrease is observed at 50% load. For a non-economizer steam genera­
tor, the feedwater condenses the steam bubbles carried into the downcomer. For
an economizer steam generator the steam bubbles are not condensed and thus the
driving density in the downcomer is less, leading to a lower recirculating flow
rate.

J-2
o
+J
ro
cm
C irc u la tio n

Based on flow area


of 12.7 ft2

Separator Shock Loss Factor

Figure J-1. "CRIBE" results Design B steam generator.

J-3
C irc u la tio n R atio

Figure J-2. "CRIBE" results Design B Steam Generator

J-4
C irc u la tio n R atio

Shock Loss Factor

Figure J-3. Shock loss factor vs. CR for economizer steam generator.

J-5
9.0

V7--0----- _

0—50% Load
~

... O ....

8.0 o—

100% Load
Circula

A—A—
4.0

3.0 .

± ±
1 2 3 4 5 6
% Carryunder

Figure J-4. Carryunder vs. CR non-economizer steam generator.

J-6
50% Load

100% Load

% Carryunder

Figure J-5. Carryunder for economizer steam generator.

J-7
REFERENCE

1. M. A. Pulich and S. G. Margolis. "CRIB1-A Steam Generator Stability


Analysis Program for the Philco-2000 Computer." WAPD-TM-530. 1965.

J-8

You might also like