United States
Environmental Protection
Agency
 Industrial Environmental Research"
 Laboratory
 Research Triangle Park NC 27711
Research and Development
 EPA-600/S7-81-019  May 1981
Project  Summary
Comparison  of  Relative
Emissions  from  Double
Mechanical  Pump  Seals to
Other  Types of  Pump
Shaft  Seals

LP. Provost, J.E. Tobias, and S.L Preston
  This study compares emissions
from double mechanical pump seals
and other types of pump shaft seals.
Analysis of emissions from the EPA
Fugitive Emissions Data Base does
not  show  significant  differences
between  double mechanical pump
seals and other types of pump shaft
seals. Both the percentage of seals
leaking  and emission  factors  for
double mechanical seals were similar
to other types of seals  in the data
base.
  This Project Summary was devel-
oped by EPA's Industrial Environ-
mental Research Laboratory, Research
Triangle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).

Introduction
  Pump seal data from nine petroleum
refineries  were extracted from  the
EPA Fugitive Emissions Data Base for
comparison of emissions from double
mechanical pump seals to other types of
pump shaft  seals.  Separate
comparisons were made for light and
heavy liquid service.
  Nonmethane leak rates for single and
double seals were  compared using
schematic plots.1 Because the emission
data have been shown to be lognormally.
distributed2, a logarithmic transforma-
tion  was  performed  before  the
schematic plots were made. The plots
illustrated that seal type had practically
no effect on leak rate for leaking seals in
light or heavy liquid service.
  Centrifugal/mechanical  pump,
centrifugal/packed  pump,  and
reciprocal/packed pump  leak  rates
were also compared with scheVnatic
plots. The plots illustrated that pump
type had little effect on non-methane
leak rate for both light liquid and heavy
liquid service.
  The percentage of seals leaking and
95 percent confidence intervals about
the percentage leaking, were calculated
to evaluate the tendency for pump shaft
seals to leak. Calculation of confidence
intervals is equivalent to performing a
Student's t-test on the groups of data.
Non-overlapping confidence  intervals
for two sets of data indicate that a t-test
would  show the averages of the two
data sets  to differ significantly.
'Tukey, John W, Exploratory Data Analysis.
Volume I, Chapter 5. Reading, Massachusetts:
Addison-Wesley Company, 1977
2Weatherold, R. and Provost, L, Emission Factors
and Frequency of Leak Occurrence for Fitting in
Refinery Process Units, EPA-600/2-79-044,
February 1979.
                                    * US GOVERNMENT PRINTING OFFICE. 1961 -757-012/7118

-------
    Similarly,  overlapping  confidence
    intervals  indicate  no  significant
    difference between the averages of the
    data sets. The overlapping confidence
    intervals illustrated that there was no
    significant difference in percentage of
    seals leaking between the single and
    double seals for either light liquid or
    heavy liquid service. There was also no
    significant difference in the percentage
    of seals  leaking  among  centrifugal/
    mechanical pumps, centrifugal/packed
    pumps, and reciprocal/packed pumps.
      The  percentage of  seals  leaking
    among centrifugal/mechanical pumps,
    centrifugal/packed  pumps,  and
    reciprocal/packed  pumps were
    compared  separately for  single  and
    double seal pumps and heavy and  light
    liquid  service.  In   all   cases,  the
    confidence intervals for the three types
    of pumps overlapped,  indicating no
    significant  difference among  pump
    types for either single or double seals
    and light or heavy liquid service.
      For a comparison ' of  both  the
    percentage  of seals  leaking and the
    amount  of leakage per seal,  emission
    factors and 95 percent confidence limits
    about  the   emission  factors were
    calculated. No emission  factors were
    calculated for cases where the number
    of pump seals screened was less than
    20 because confidence limits for such
    small sample  sizes would have been
    extremely wide. For this reason, there
    was little information for comparison of
    the effect of pump seals on  emission
    factors for any particular service and
    seal type. Where enough information
    for  comparisons existed, all  emission
    factor confidence intervals overlapped.
    This  indicated that  there  was  no
    significant   difference  in   emission
         factors between single and double seals
         for either light or heavy liquid service,
         no significant difference in emission
         factors among centrifugal/mechanical,
         centrifugal/packed,  and  reciprocal/
         packed pumps for either light or heavy
         liquid  service,  and   no  significant
         difference between   centrifugal/
         mechanical  and  centrifugal/packed
         pumps for single seal pumps and heavy
         liquid service.

         Conclusions
          Analysis  of  emissions  from  the
         refinery data base shows no significant
         differences between mechanical pump
seals  and other  types of pump shaft
seals.

Recommendations
  Lack of significant differences maybe
due  to  the  small  amount of  data
available when subcategories of pump
seals were analyzed. Also, the effect of
other  variables  such   as  point  of
emission measurement,whether or not
the  pump  was  running,  and   the
particulars of the service for each pump,
may mask differences in performance of
the various seal types. A more complete
study should consider  these variables
and be based on  a larger sample size.
           L P. Provost, J. E. Tobias, and S. L. Preston are with Radian Corporation, Austin,
             TX 78766.
           S. L. Rakes is the EPA Project Officer (see below).
           The complete report, entitled "Comparison of Relative Emissions from Double
             Mechanical Pump Seals to Other Types of Pump Shaft Seals," (Order No.
             PB 81-162 083; Cost: $5.00, subject to change) will be available only from:
                   National Technical Information Service
                   5285 Port Royal Road
                   Springfield,  VA 22161
                   Telephone: 703-487-4650
           The EPA Project Officer can be contacted at:
                   Industrial Environmental Research Laboratory
                   U.S. Environmental Protection Agency
                   Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
               Postage and
               Fees Paid
               Environmental
               Protection
               Agency
               EPA 335
Official Business
Penalty for Private Use $300
    PS   0000329
    U  S  ENVTR  FPOTt'CXIUN  AGFINCX
    KEGION  5  LiBrtAKY
    230  S  DEARBORN STREET
    CHICAGO  IL 60604

-------