United States
Environmental Protection
Agency
Industrial Environmental
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-84-047 May 1984
Project Summary
Long-Term Continuous Monitor
Demonstration Program:
Columbus and Southern Ohio
Electric Company, Conesville
Unite
Edward F. Peduto, Jr.
A continuous emission monitoring
(CEM) demonstration program was
conducted at the Columbus and South-
ern Ohio Electric Co.'s Conesville
Generating Station. The primary pur-
pose of this program was to demonstrate
the feasibility of the monitoring require-
ments specified in 40 CFR, Part 60,
Subpart Da. A secondary objective was
to adhere to the draft quality assurance
requirements scheduled for promulga-
tion as Appendix F.
An assessment of system availability
as a percentage of hourly unit uptime
indicated an average of 78 and 75
percent availabilities for the SO2 outlet
emissions and control system removal
efficiency reporting parameters, respec-
tively. Availability for the NOx system
averaged 65 percent. Subpart Da
availabilities (number of valid 30-day
rolling averages) were 67, 59, and 62
percent for the SOz emissions, efficiency
and NOx emissions reporting param-
eters, respectively.
Evaluation of labor requirements
indicates that the total average weekly
level of effort was 27 hours. Of the
total, 64 percent are attributable to
dairy operations, 23 percent to nonrou-
tine maintenance, and the remaining 13
percent to preventive maintenance.
The system consistently complied
with the performance specifications
criteria. Relative accuracy results
were well within the 20 percent limit
except on one occasion when the NOx
result slightly exceeded this level.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering
information at back).
Introduction
On June 11,1979, the U.S. EPA promul-
gated new source performance standards
(NSPS) for new utility steam generators
for which construction commenced after
September 18, 1978. To demonstrate
compliance, EPA also promulgated a
reference method (Method 19) that
supplies a protocol for determining the
control device, sulfur input rate by either
fuel sampling and analysis or continuous
monitors, and the final sulfur emissions
to the atmosphere by continuous flue gas
analysis. In addition, the sulfur removal
efficiency for the control device is
calculated using these measurements.
This method was used to collect data for
developing the current set of performance
standards.
EPA's Office of Air Quality Planning
and Standards, assisted by the Technical
Support Office of EPA's Industrial Envi-
ronmental Research Laboratory at Re-
search Triangle Park (IERL-RTP), initiated
a program to demonstrate the feasibility
-------
of the requirements for monitoring and
control of SOa emissions. This program
focused on sources subject to the NSPS
specified in 40 CFR 60, Subpart Da.
GCA/Technology Division was con-
tracted to conduct the portion of the
program associated with measuring
gaseous emissions. The primary objective
of the program was to demonstrate the
feasibility of the monitoring requirements
specified in Subpart Da. Of secondary
importance was the development of
background and support data for quality
assurance regulations. The study involved
a planning and procurement stage, a 1-
year field demonstration phase, and final
evaluation of the data. Approximately 12
months of data were collected during the
field portion of the program.
General Program Approach
The approach simulated activities
involved in preparing for and complying
with the Subpart Da requirements. To
fulfill these requirements: the utility must
prepare a monitoring system specification;
procure, install, and certify the system;
and (finally) operate the system. Subse-
quently, the data are reported according
to the Subpart Da specifications. In this
program, a number of prospective utility
sites were surveyed as possible test
locations. Since no sources subject to
Subpart Da were operational, newer
operational facilities subject to Subpart D
specifications were considered. The test
site selection criteria were:
• Coal-fired steam generator, 100
MWE minimum, less than 10 years
old.
• Coal sulfur content, 2-5 percent;
and medium to high ash content.
• Flue gas desulfurization (FGD)
system designed for a minimum of
80 percent SO2 removal on 100
percent of the flue gas.
• FGD system which employs multiple
scrubber modules which are parallel.
Installation of the monitoring system
was followed by a start-up and trouble-
shooting phase, during which the system
was operated and debugged. Shortly
thereafter, the system was certified
according to proposed Appendix B "Per-
formance Specifications Tests 2 and 3,"
specified in the October 10, 1979,
Federal Register. After certification, the
1 -year field demonstration involved:
• Routine operation according to
Method 19 and Proposed Quality
Assurance Requirements.
• Reporting consistent with Subpart
Da.
• Data Collection and Quality Assur-
ance activities for assessing moni-
toring economics and system per-
formance.
Routine activities were conducted daily
by an onsite technician. Additional data
were collected during quarterly quality
assurance and stratification tests.
After the test phase, all data were
evaluated, including an assessment of
continuous emission monitor (CEM)
availability and data capture, the costs of
procurement and operation, and monitor-
ing system performance. System perform-
ance is discussed in reference to 40 CFR
60, Appendix B, and proposed quality
assurance protocols.
Test Facility Description
The Conesville Generating Station's
Unit 6 was selected as the site for the
CEM demonstration program. In Unit 6,
the flue gas exiting the boiler passes
through two, parallel, air preheaters and
into two, parallel, electrostatic precipitator
modules. The gas from the precipitators
passes into two, double-inlet/single-
outlet, ID fans, which move the gas
through the two scrubber modules. The
modules are spray types designed to
operate with free floating balls; however,
the balls were removed to reduce module
pressure drop and increase flow capacity.
Each module can handle 50 percent of
the flue gas at full load conditions. Gas
exiting the scrubber modules and the
' bypass duct are combined into a single
common duct before entering the 244 m
(800 ft) tall stack.
At full generating capacity, total flue
gas flow rate through the system is about
1.3 x 106 acfm (36.8 x 103 mVmin) at
300°F (149°C). About 30 percent of the
total flue gas flow is bypassed for reheat
purposes around the scrubber modules
during full load conditions.
During the initial period of the Field
Demonstration Phase, Unit 6 was oper-
ated year-round as a baseload unit. Unit 6
normally operated at full-load capacity
during daytime peak power demand
periods; at night, the unit operated at
half-load with only one scrubber module
in service. During the later stages of the
demonstration period, Unit 6 operated
intermittently as a baseload unit and was
frequently taken offline for varying
lengths of time as a result of the low
power demands during spring and early
summer of 1982. The unit was never
operated in a swing load configuration.
Continuous Emission
Monitoring System (CEMS)
Description
The overall design objectives for the
CEMS were to meet Subpart Da monitor-
ing and reporting requirements while
minimizing manpower requirements. To
meet these requirements, the following
CEM capabilities were required:
• Dual inlet FGD emissions monitor-
ing for S02.
• Outlet FGD emissions monitoring
for S02 and NOX.
• Daily emissions and removal effi-
ciency average calculations.
• 30-Day rolling averaged emissions
and removal efficiency calculations.
These requirements were met with a
system approach entailing automatic
CEMS operation.
Monitoring points were at each scrub-
ber module inlet and at the combined
outlet breaching prior to entering the
stack. At the inlet to each scrubber
module, a single sample probe was
downstream of the ID fans and just
upstream of a guillotine bypass damper
door. A single outlet sampling point was
in the common breaching, downstream
from both the scrubber module and
bypass duct outlets.
The multilocation extractive system
was specified by GCA and designed and
constructed by KVB, Inc., of Irvine, CA. The
system consisted of two sets of gas
sensors and the conditioning hardware
and data acquisition equipment listed in
Table 1. One set of instruments measured
inlet SOz and 02 on a time-shared cycling
basis, and a second set measured outlet
concentrations of SO2, 02, and NO*.
Samples were conditioned by filtering
particles using an in-stack filter and a
high-surface area, glass fiber filter, both
at the extraction point. A dual-pass
condensation unit in the instrumentation
shelter removed moisture. The KVB
system operated automatically, including
multipoint calibration, zero/span checks,
automatic data acquisition, and onsite
data reporting.
Routine Operation
During the field demonstration period,
a CEMS operator was onsite most of the
time Unit 6 was online. The operator
performed dialy diagnostic checks,
maintenance, and logistical support
services for the CEMS. In keeping with
the approach of minimal operator atten-
tion, the CEMS operator left the site and
checked the CEMS status remotely using
the modem link during Unit 6 outages.
The modem link was also used when an
operator was not onsite when Unit 6
was operating.
Daily, weekly, and monthly operator
routines involved CEMS diagnostic
checks, checking analyzer calibrations,
generating daily reports, performing
-------
Table 1. Major KVB System Components
INLET MEASUREMENTS
• SO2 — DuPont Model 400; 0-5000 ppm
Oz — Thermox WDG III, 0-25 percent
OUTLET MEASUREMENTS
• SO2 — DuPont Model 400; 0-1000 ppm (Normal Range 100O-5OOO ppm (Auto range)
/VOx — Thermo Electron Model 10A; 0- 1OOO ppm
Oz — Thermox WDG III; 0-25 percent
COMPUTER SYSTEM
• Cromemco System Three
• Dual 8 in. floppy disc drive
• CRT Terminal
• Report Format Printer
EXTRACTION SYSTEM
• Paniculate Removal—Filtration
• Moisture Removal—Dual Pass Condensation
routine and nonroutine maintenance,
computer operations, and general admin-
istrative duties, including project docu-
mentation and logistical support.
The CEMS diagnostic checks were
performed twice daily, including visual
checks of various CEMS operating
parameters such as sample flow rates,
vacuum and pressure levels, and temper-
atures, to ensure that all were within
specified operating limits.
Calibration checks were controlled
automatically by the computer at midnight,
and the results were evaluated by the
operator each morning. If calibration
results indicated excessive drift, the
operator terminated CEMS operations
and manually recalibrated the affected
analyzer.
The CEMS operator reviewed the
process and emissions data, and noted
any process upsets (e.g., scrubber
malfunctions) and any emissions data
invalidations (e.g., periods when an
analyzer was offline for repair or an
analyzer was malfunctioning). The time
and nature of process upsets or data
invalidations were recorded in the CEMS
uptime log.
Discussion of Results
The KVB system was operated and
performance tested over a period of 16
months in which Unit 6 was operational
for approximately 12 months. During this
period, various external quality assurance
tests were conducted in addition to the
routine monitoring activities. In total,
program activities provided the necessary
data for assessing system performance,
system availability, and monitoring
economics. The following subsections
summarize each area.
Assessment of Monitoring
System Performance
The continuous monitoring system was
operated and maintained according to 40
CFR 60, proposed Appendix B. In addition,
other tests were conducted to further
evaluate system performance, including
initial and final performance specifica-
tions tests, stratification testing, relative
accuracy audits, and calculation of
precision estimates.
Stratification Tests
Before initiating the Operational Test
Period and the routine monitoring phase,
stratification tests were conducted at the
monitoring locations, to ensure that the
probes were placed to provide represent-
ative flue gas samples. The procedure
consisted of measuring the SOz and Oa
concentrations at each designated tra-
verse point. Between each pair of ports
(three traverse points per port), the
monitoring system was switched from
the traverse point to a reference point.
Data from the reference point were used
to correct for temporal process variations
during the test. Temporally corrected
traverse data were used to define the
spatial variability of the flue gas stream.
One stratification test was conducted
at one of the two inlets. Additional tests
were not conducted because the probabil-
ity of stratification is low at scrubber
inlets.
During the program, eight stratification
tests were conducted at the outlet: four at
full load with both modules online, and
the remaining four with the boiler
operating at half load and one module
online. Two such tests were conducted
for each module.
Performance Specifications
Tests
Monitoring system performance was
assessed using the Performance Specifi-
cations Tests outlined in the October 10,
1979, Federal Register. These tests
provide a basis for determining adherence
to minimum compliance monitoring
system performance levels. Tests were
conducted to quantify both short- and
long-term drift, calibration error (preci-
sion), response t i me, a nd system accu racy
relative to the reference methods.
These tests were conducted during the
latter part of May and the first week of
June 1981. An abbreviated version of
these tests was conducted during the last
month of the program (August 1982).
During this last test session, the relative
accuracy and calibration error tests were
repeated.
The entire monitor system conformed
to all performance criteria, except that the
outlet SOa monitor exceeded the 24-hour
drift and mid-scale calibration error
criteria. As a result of the drift exceedance,
the monitor was equipped with an
automatic zero function which rezeroed
the analyzer every 15 minutes. The mid-
scale calibration error result exceeded
the specification during the first tests,
but was within limits during the final
tests.
The relative accuracy tests conducted
at the beginning and the conclusion of the
demonstration phase indicate perform-
ance consistent with regulatory require-
ments. During the initial tests, the inlet
and outlet SOz system relative accuracies
(emission rate) were 6.5 and 15.6
percent, respectively. Results for the
concluding test were comparable with
inlet and outlet relative accuracies of 9.8
and 4.8 percent, respectively. Emission
rate relative accuracies for the NOX
monitoring system were not appreciably
different when comparing the two tests.
Initially, the relative accuracy was 15.2
percent; the concluding result was 14.8
percent.
Accuracy and Precision
As an extension of current monitoring
methodologies, EPA's Quality Assurance
Division is formulating and evaluating
quality assurance measures applicable to
power plant monitoring. These procedures
include protocols for demonstrating the
accuracy of data relative to a reference
method, as well as a simple daily routine
to determine the necessary data for
assessing the precision of the measure-
ment equipment on a monthly basis.
To demonstrate the utility of these
methods, GCA conducted the accuracy
and precision estimating methods. Various
draft versions of Appendix F were
prepared during the 1980-1982 time
frame. The precision and accuracy
checks were conducted according to the
November 19, 1981, version.
-------
Accuracy
Proposed Appendix F provides two
options for assessing system accuracy:
one involves standard gases in which the
standard gas is analyzed and processed
as a flue gas sample; and the other
involves side-by-side comparisons of
CEMS outputs using applicable reference
methods. For this program, the reference
method was chosen.
Results of the relative accuracy audits
are shown in Table 2, which includes
results of the initial and final relative
accuracy tests. Each relative accuracy
audit consisted of at least six valid test
replicates; whereas, the initial and final
relative accuracy tests consisted of nine
replicates.
The inlet and outlet S02/O2 moaitoring
system exhibited relative accuracies in
the range of 4.8 to 16.9 percent on an
emission rate basis. When reviewing
these results, note that the monitoring
system and reference methods each
contributed an imprecision and bias
component to the relative accuracy
result.
N0« results were in the range of 9.7 to
23.2 percent on an emission rate basis.
During the July 1982 audit, the relative
accuracy exceeded the allowable 20.0
percent criterion. This result indicates a
comparable bias to other test periods;
however, the confidence interval (preci-
sion component) was much larger than
before.
Precision
Data from the daily calibration checks
were used to estimate the upper and
lower probability limits for each analyzer.
These estimates were determined using
the protocol in the proposed Draft
Appendix F Quality Assurance procedures,
dated November 19, 1981.
Daily, the zero and high span gases
were input through the entire system
(including all extraction equipment) and
analyzed as a sample gas. The resulting
voltages were substituted in the appropri-
ate calibration equation to determine the
corresponding concentration in units of
ppm S02, ppm IMOx, or percent 02. Using
these data as inputs, the upper and lower
probability estimates for the zero and
high span levels were calculated by
determining the relative percent differ-
ences between the measured span (zero)
gas concentration and the accepted
concentration.
CEMS Availability
For this discussion, two definitions
of availability are presented. Total
Table 2. Summary of Relative A ccuracy A udits
Date
June 1981
July 1981
June 1982
July 1982
August 1982
S02
4.3
11.6
6.5
7.4
9.8
Met
Oi
0.46
0.55
0.68
1.53
0.53
Outlet"
System
6.5
15.0
7.8
9.4
9.8
SO2
10.9
16.0
13.3
12.9
3.9
02
0.28
0.40
0.19
0.24
0.43
System
15.6
16.9
12.8
14.2
4.8
Outlet"
NO*
12.8
6.7
17.2
23.7*
12.7
System
15.2
9.7
17.7
23.2"
14.8
"SOa /VO,, and system are in terms of % relative accuracy; O2 is in terms of % Oz concentration.
"Exceeded the allowable criterion of <2O%.
availabilities refer to a simple proportion
of CEMS uptime to boiler uptime on an
hourly basis. This proportion is also
modified to conform to Subpart Da
requirements, which involve definitions
of valid periods of data that are required
for reporting purposes. A valid data hour
must consist of at least two quarters of
data; and a valid daily average must
contain at least 18 hours of data for a 24-
hour, midnight to midnight, boiler operat-
ing day. A valid 30-day rolling average
must contain at least 22 valid days of
data.
Figure 1 shows valid monitoring data
availability based on the number of
boiler operating hours in each calendar
month. The primary Subpart Da reporting
parameters are outlet SO2 and NOX
emission rates (ng/J) and the control
unit S02 removal efficiency. Emission
rates are calculated using the data
generated by the pollutant and diluent
monitor combination. Similarly, the SOz
removal efficiency is calculated using
concurrent data generated by the inlet
WOf
SO2/O2 and outlet SO2/02 monitor
pairs.
The hourly availability rates shown in
Figure 1 range from 38 to 90 percent for
the outlet S(?2 emission rate, 18 to 90
percent for the NOX emission rate, and 35
to 88 percent for removal efficiency.
Overall, the outlet SO2 emission rate
availability averaged 78 percent; the NOX
emission rate, 65 percent; and removal
efficiency, 75 percent.
Table 3 summarizes the overall availa-
bility, indicates the adherence to the
Subpart Da data capture requirement on
a daily basis, and shows the data
availability for reporting the 30-day
rolling averaged data. In total, there were
277 "boiler operating days (as defined in
Subpart Da)" during the demonstration
phase. For these boiler operating days,
the daily data capture rate for outlet S02
emission rate was met 72 percent of the
time; the removal efficiency availability
requirement was slightly less (69 percent).
The daily NOX reporting requirement was
met 59 percent of the time. About half of
i I I I l I I I I I i r
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug
m 1981 9*"^ 1982 *•
Figure 1. Monitoring data availability by calendar month for outlet SO2 and NO „ emission rates
and S02 removal efficiency (percentage of CEMS uptime relative to total boiler
uptime).
-------
Table3. Availability Summary for Required Subpart Da Reporting Requirements
Parameter units
Outlet
SO2
Emissions
Outlet
NO,
Emissions
Removal
Efficiency
Total availability
% total Unit 6 uptime
24 hour availability*
% valid 18-24 hour days
Subpart Da availability
% of valid 30-day rolling averages
78
72
67
65
59
62
75
69
59
"Based on 277 total boiler operating days.
the NOx data unavailability was due to
NOx vacuum pump failure.
Subpart Da 30-day rolling average
availabilities were 67 and 59 percent for
the SOz outlet emissions and removal
efficiency, respectively. The availability
for the NOX reporting requirement was 62
percent. For rolling averages that do not
fulfill the minimum data capture require-
ments, the upper and lower confidence
intervals for each parameter are calcula-
ted. The upper limit is used for reporting
outlet emission rates, and the ratio of the
upper outlet emission limit and the lower
inlet interval limit are used to report the
efficiency parameter.
Assessment of Economic
Requirements
The economic impact of continuous
monitoring regulations is also an area of
concern. The cost data presented are
based solely on actual costs associated
with the Conesvilie program. The applica-
bility of these costs to other utilities
installations provides a basis for projec-
tions. Proper use of the experience gained
in this program should significantly
reduce certain costs; however, site-
specific variables must be considered if
meaningful projections are to be obtained.
The utility operator experiences two
cost categories in fulfilling the Subpart
Da reporting requirements. Initially,
major capital expenditures are made in
purchasing, installing, and certifying a
monitoring system. Subsequently, ongo-
ing costs result from routine operation
and data reporting.
The total $269,030 "premonitoring"
costs for the Conesville CEMS include: (1)
the system purchase price, $181,000; (2)
the cost to build a temperature controlled
monitor room and to route electrical
wiring and sample line, $31,030; (3) final
installation, start-up, and checkout,
nearly $36,000; and (4) the certification
tests, $21,000.
Operating costs at Conesville are
shown in Table 4. The 1,438 labor hours
include 915 (68 hr/mo) for daily opera-
tions, 327 (24 hr/mo) for nonroutine
maintenance, and 196 (16 hr/mo) for
preventive maintenance. Nonroutine
material costs were $2,769, routine
expendable material costs were $2,869,
and calibration gas costs were $8,607.
(Note that the nonroutine material costs
are biased low because many of the
replacement parts were supplied under
warranty, free of charge.)
A comparison of the CEMS operating
costs with the total quality assurance
activities is given in the full report. The
quality assurance costs were $90,000
and covered an initial and final perform-
ance specifications test series, three
abbreviated relative accuracy audits, four
stratification test series, and the calcula-
tion of monthly precision estimates. The
utility operator is required to conduct the
initial performance specifications test
series and a location representativeness
(stratification) test. Future regulations
may include the reporting of precision
estimates and the conduct of a periodic
relative accuracy audit and/or a cylinder
gas audit.
Conclusions
While program results do not fully
answer all questions concerning contin-
uous monitoring, several conclusions
and/or judgements can be made. State-
ments concerning this program should be
prefaced by several qualifiers. First and
foremost is that the experience and
results may be specific to the Conesville
monitoring situation. Second, several
economic constraints precluded imple-
mentation of system modifications which
would have most likely improved system
uptime.
Design and Software
Deficiencies
Several design deficiencies noted with
the CEMS severely hampered availability
and the subsequent fulfillment of moni-
toring requirements. The primary prob-
lems were not hardware related, but
more "action and reaction" related;
specifically, system reactions to alarm
conditions. KVB configured this system to
be totally automated: all operation and
control functions are "slaved" to the
computer. As a result, normal data
acquisition ceased whenever computer
operations terminated. Provisions should
be made for the system to be operated
manually during computer outage.
Other problems were related to the
conditioning and switching system.
Responsible for most of the hardware-
related downtime was the valve switching
system, which primarily used solenoid
valves which are prone to sticking. These
could be replaced with multiport ball
valves.
The NOx converter and vacuum system
were responsible for most of the NOX
downtime. When exposed to high-sulfur
and -chloride laden gases, the stainless
steel converters show a high corrosion
rate. Bypassing the converter would
eliminate this problem. The NOx vacuum
pump was failure prone for unknown
reasons: it appears that the best approach
for maintaining uptime is to keep a spare
pump on hand.
The computer disc system was very
failure prone. Repeated read/write errors
resulted in significant data loss. Much of
the problem can be attributed to dust.
Aside from the read/write problems,
the durability of the floppy disc system for
field applications is questioned. During
the 16 month demonstration phase, the
disc system was replaced once and
underwent significant component re-
placement on another occasion. For
future applications, a sealed Winchester
or electronic disc storage device is
recommended.
Availability
Overall availability for future systems
of this sort can be improved significantly
by implementing several generic design
changes. Based on the results of this
study, future systems should be configured
in two basic subsystems: the condition-
ing/monitoring subsystem and the data
storage/reduction system. In addition,
these two systems should be separable;
i.e., if the data system is removed, the
monitoring subsystem should continue to
collect and store data on a secondary
system such as a secondary buffer or strip
chart recorder. Data availability should be
significantly higher since the monitoring
process is not terminated upon the
termination of computer control.
Other considerations should be given
to the servicing of alarm conditions.
Often, the computer system terminates
system operation for what may be a very
minor problem. The power plant environ-
-------
Table 4. OEMS Operation Manhour Requirement Summary
OEMS Maintenance Summary
Total
Hours
% Total
Weekly
Average
Hours
Conditioning
System
NR" R"
242 62
17% 4%
4.5 1
Analyzers Computer
NR
60
4%
1
R NR
118 25
8% 2%
2.2 0.5
ft
re
1%
0.3
Total
NR
327
23%
6
R
196
14%
4
Daily
Operations
915
63%
17
Total
Manhour
Requirements
1438
27
*NR = Nonroutine Maintenance.
b/? = Routine Maintenance.
ment is well geared to the use of
enunciator panels to report alarm condi-
tions. After acknowledging these types of
alarms, the appropriate action is taken by
operators or technicians.
Automatic routines such as calibrations
should be scheduled during time periods
when more personnel are on duty.
Considerable downtime was encountered
during the Conesville program when an
automatic function occurred at midnight,
activating an alarm. The alarm would not
be discovered until morning because
personnel were unavailable to correct the
problem.
Projected Utility Costs
The operational experience gained
during the program can be used to
determine projected costs to the utility
industry. Yearly operations comprise the
largest cost element associated with the
monitoring. An estimated 900 manhours
per year will be expended to operate the
system, verify the data being stored, and
acquire the proper process and monitor
documentation for preparing the quarterly
Subpart Da and Appendix F reports.
Quarterly relative accuracy audits
constitute the next major labor require-
ment. Approximately 140 manhours per
audit will be expended if conducted by
inhouse personnel. An outside contractor
would charge approximately $10,000 per
audit including travel.
Routine maintenance labor constitutes
about 8 percent of the yearly level of
effort, or 210 manhours. This level may
be elevated slightly as the maintenance
schedules are refined. The nonroutine
maintenance level should be reduced
with a slight increase in preventive
maintenance.
Nonroutine maintenance has been
estimated at slightly over 310 manhours
per year. This is higher than expected, but
should decrease with time as the preven-
tive maintenance program is refined.
Quarterly Subpart Da and Appendix F
(Draft) reports are projected at a combined
yearly total of 220 manhours, based on
using an automated data reduction
system.
Calibration gas cost is the single largest
quarterly expendable material expense
and most predictable utilizing automatic
analyzer calibration systems in which the
gas is injected at the probes. The
estimated yearly cost (based on the
Conesville system) is $13,940 using
Protocol I certified span gases. Other
types of systems which use gas diluters
for blending the various span gas
concentrations will require about 30
percent of the gases required for Cones-
ville.
The spare parts inventory cost is a
somewhat significant addition to the
original purchase cost of a CEMS. Usage
rate and quarterly replacement cost are
difficult to predict. A good warranty
maintenance agreement with the CEMS
vendor can drastically reduce this cost for
the first years of CEMS operation. An
adequate spare parts inventory for the
Conesville system would approach $11,600,
or 6.4 percent of the system cost.
i
Edward F. Peduto, Jr.. Timothy J. Porter, and David P. Midgley are with GCA/
Technology Division. Bedford. MA 01730.
D. Bruce Harris is the EPA Project Officer (see below).
The complete report, entitled "Long-Term Continuous Monitor Demonstration
Program: Columbus and Southern Ohio Electric Company, Conesville Unit 6,"
(Order No. PB 84-178 649; Cost: $13.00, 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
Official Business
Penalty for Private Use $300
* U.S. GOVERNMENT PRINTING OFFICE: 1964-759-102/950
------- |