EMB Project No. 77-SPP-21A
CD
•^
o
POLLUTION
EMISSION TEST
FINAL REPORT:
OPERATION OF A CONTINUOUS SULFUR DIOXIDE
MONITORING SYSTEM
AT
.-.•A-^a^ai-.i •.•••.•.-.• A^iJ CANE RUN UN IT 4
LOUISVILLE GAS & ELECTRIC CO.
LOUISVILLE, KENTUCKY
JULY 1978
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Emission Measurement Branch
Research Triangle Park. North Carolina
-------
EMB Project No. 77-SPP-21A
SET 1677 01 0678
FINAL REPORT:
OPERATION OF A CONTINUOUS SULFUR DIOXIDE
MONITORING SYSTEM
AT
CANE RUN UNIT 4
LOUISVILLE GAS & ELECTRIC CO.
LOUISVILLE, KENTUCKY
Prepared For;
U, S, Environmental Protection Agency
Emission Measurement Branch
Research Triangle Park, NC 27711
July 1978
Prepared By:
Mark Repp
SCOTT ENVIRONMENTAL TECHNOLOGY, INC.
Plumsteadville, Pennsylvania 18949
Scott Environmental Technotosy
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
2.0 SUMMARY 3
2.1 SYSTEM AVAILABILITY 3
2.2 S02 MONITOR - OPERATIONAL SUMMARY . . . 3
2.3 02 MONITORING SYSTEM - OPERATIONAL SUMMARY . 5
3.0 DESCRIPTION OF MONITORING SYSTEM . . . 8
3.1 DUPONT 400/460 S02 MONITORING SYSTEM . 8
3.2 SCOTT MODEL 150 02 MONITOR 10
3.3 02/S02 SYSTEM 10
4.0 OPERATING PROCEDURE 15
4.1 CALIBRATION 15
4.2 DATA COLLECTION AND REDUCTION 16
5.0 DISCUSSION OF RESULTS . . 17
5.1 ROUTINE MAINTENANCE 17
5.2 NON-ROUTINE MAINTENANCE 18
6.0 RECOMMENDATIONS 21
Scott Environmental Technofosy Inc
-------
SET 1677 01 0678
1.0 INTRODUCTION
This report summarizes the operation and maintenance of the
continuous SOo/Oo monitoring program conducted at Louisville Gas & Electric
Company's Cane Run Unit No. 4. The monitoring program was conducted by
Scott Environmental Technology, Inc. under contract to the Environmental
Protection Agency (Contract No. 68-02-2813). The program ran from July 21,
1977 through April 13, 1978 and included the initial performance testing
of both S02 and ©2 monitors. These performance test results are discussed
in a report entitled "First Interim Report on Continuous Sulfur Dioxide
Monitoring Program at Cane Run Unit No. 4". This report, however, only
considers the operation after August 5, 1977 when the oxygen system was
installed. Also, the period from January 24 to April 4, 1978 is not
considered because the scrubber was shut down and the boiler ran under
abnormal conditions due to severe winter weather and a coal strike.
The scope of the program called for the operation of both S02
and 02 continuous monitors at two inlets and two outlets of a dual module
mobile bed carbide lime scrubber.
The S0~ monitor was a DuPont Model 400/460 system which included
sample lines from each of the four sample locations and a switching net-
work that sequentially sampled from each location every 15 minutes. The
oxygen monitor was a Scott Model 150. Sample for the 02 monitor was taken
from the DuPont sampling system to enable monitoring at four locations
using a single 02 instrument.
All strip chart data and operation logs were sent to Monsanto
Research Corporation for reduction and computer analysis. Weekly site
visits were made by Scott personnel to maintain the instruments and collect
data. However, on November 28th the scope of work was changed so that
Scott sent reduced data to Monsanto rather than the raw strip chart records,
Also, Scott personnel remained on site Monday through Friday of each week.
This arrangement was continued through the end of the program.
This report considers only the operation and maintenance of the
S02/02 system and does not deal directly with any S02 or 02 data. Included
Scott Environmental TechnoSosy Inc
-------
-2-
SET 1677 01 0678
are data on system availability, system design, maintenance requirements
and recommendations based on experience gained during this program.
Scott Environmental "fechnotogy Inc
-------
-3-
SET 1677 01 0678
2.0 SUMMARY
2.1 SYSTEM AVAILABILITY
System availability is defined as the amount of time that an
individual system component was functioning and producing valid data
expressed as a percentage of time the scrubber was operating. The monitor-
ing program encompassed 4344 hours from August 5, 1977 to April 13, 1978,
not including the January 24-April 4 scrubber outage. During the program
the boiler was off line for 534 hours. Of the remaining 3810 hours, the
scrubber was on line 3512 hours producing a total scurbber availability
of over 92%. Calculations were not performed to express the instrument
system availability as a function of boiler operating time. If it is
assumed that the instrument systems were out of service during scrubber
outages, the availabilities presented below would be approximately a factor
of 0.92 lower. However, the usual case was that the instrument systems
were operational during periods of scrubber outage. It is estimated that
availabilities based on boiler operating time would not be significantly
different than those presented below which are based on scrubber operating
time.
The availability of the SC^ and 62 monitors is shown in Table 1.
These figures are based on the 3512 hours that the scrubber was in service.
In the case of the S02 monitor, the table shows the availability of two,
three and four sample locations. Situations with only two or three sample
locations in service were due to clogged probes or other problems with
individual sampling systems. This individual sample location breakdown
was not necessary with the 62 monitoring system because it employed the
SC>2 sampling system and individual sample point failure could only be
attributed to the DuPont S02 system. Therefore the 02 system was con-
sidered available if it produced data from any sampling points.
Scott Environmental Technctogy Inc
-------
-4-
SET 1677 01 0678
TABLE 1
S02/02 SYSTEM COMPONENT AVAILABILITY*
System
Component
DuPont S02 4 points
DuPont S02 3 points
DuPont S02 2 points
02 system w/initial Scott 150
02 system after 11/17
w/replacement Scott 150
and system modifications
Total Time
Available
Hours
2589
3256
3479
1462
1318
Out of
1596 hours
Availability
74
93
99
42
83
*Calculations based on 3512 hours of scrubber operation.
Scott Environmental Technology Inc
-------
-5-
SET 1677 01 0678
2.2 S02 MONITOR - OPERATIONAL SUMMARY
Consideration of the operation of the DuPont 400/460 S02 monitoring
system at Louisville Gas & Electric must be broken down into the performance
of the monitor itself and that of its integral sampling system. The data in
Table 1 shows that the DuPont Model 400 basic photometric analyzer which
includes the instruments electronics and detector mechanism performed
extremely well. This is evident by its 99% availability during the program.
Considering the normal maintenance that is required, this figure could not
be improved significantly.
The DuPont sampling system, designated as the Model 460 four
point system, which includes sample line heaters, sample lines, probes and
backpurge mechanism, was, however, prone to failure. This is shown in
Table 1 by the lower availability of all four or even three sampling legs.
The primary problem resulting in 1 or 2 sample locations being out of
service was clogged probes. On a few occasions this was the result of
freezing of condensed water between the probe and the heated portion of the
sample line. This probe blockage occurred mostly on the inlet probes where
high pressure purge mechanisms were not installed. These high pressure
purge lines were installed on both outlet probes and probe blockage at
these locations was not a chronic problem. Therefore if these high pressure
purge lines were installed at both inlet probes the problem of probe blockage
would be significantly reduced.
2.3 02 MONITORING SYSTEM - OPERATIONAL SUMMARY
The rather low overall availability figure for the oxygen monitor-
ing system resulted from several maintenance and system application problems.
System application problems were those associated with the combining of the
DuPont sampling system with the Scott 0~ system. The maintenance data
behind the availability calculations for the 0,, system, however^ includes
two major periods of Oo outage which comprise 70% of the total Oo system
down time. These occurred from 8/17 to 9/21 and from 9/29 to 11/17. The
Scott Environmental Technotosylnc
-------
-6-
SET 1677 01 0678
8/17-9/21 outage was the result of both 0~ monitor malfunction and system
application problems. During this period site visits were on a weekly
basis and due to the unreliability of the Q£ system, little valid data was
collected. The 9/29 to 11/17 outage encompassed a period where attempts
were made to obtain an alternate 02 monitor due to chronic reliability
problems with the existing system. However, a literature review of several
alternate monitors with reasonable delivery schedules showed that many of
the original applications problems would still remain. Therefore it was
decided to redesign the (^ monitor system using a replacement Scott Model
150 oxygen analyzer.
On November 17, a modified system was installed that worked
adequately for the remainder of the program. The modifications, in addi-
tion to replacing the 02 monitor, included moving the 02 system inside.'the.
boiler house and the use of a refrigerated coil condenser to dry the sample
prior to the ©2 monitor. From November 17 to April 13th, the end of the
monitoring program, the modified 0_ system availability was 83%.
Beginning with the end of September, discrepancies were noted
between the oxygen levels measured at the outlet ducts using a Fyrite and
those indicated by the 02 monitor. These discrepancies were intermittant
and at times were as much as 4-5%. Due to the intermittant nature of the
problem isolation of the cause was difficult. Initially, through a process
of elimination, the cause was attributed to leaks in the high pressure
blow-back valves located at the ducts. Replacement valves were installed
but the problem remained. Further investigation revealed that if a proper
balance between the DuPont flow rate and that of the 02 system were not
maintained that ambient air could be drawn in from the DuPont auxiliary
air regulator and thereby produce erroneous 02 readings. Subsequent
system checks proved that this was not occurring.
During the first week in January, the cause of the problem was
isolated. The scrubber bypass damper on the south outlet duct was jammed in a
partially closed position. This damper was located immediately downstream of the
Scott Environmental Technology Inc.
-------
-7-
SET 1677 01 0678
monitoring system probes. This created a null flow zone in the area before the
damper where the probe was located. The problem was remedied by extend-
ing the probe another five feet, into the mainstream of gas flow. The
same situation existed at the north outlet because the damper had to be
similarly closed to maintain system balance. The intermittant nature of
the problem was related to varying boiler loads and subsequent variations
in duct flow. When duct flows and boiler loads were low the discrepancies
would be most severe.
Scott Environmental Technotogy !nc
-------
-8-
SET 1677 01 0678
3.0 DESCRIPTION OF MONITORING SYSTEM
The monitoring system employed at Louisville Gas & Electric
Company's Cane Run Unit No. 4 provided for the semi-continuous monitoring
and recording of both S02 concentrations and oxygen levels. These parameters
were monitored at two inlets and two outlets of a dual module carbide lime
scrubber. Figure 1 is a diagram of the scrubber system and sample probe
locations.
S0» concentrations were monitored with DuPont Model 400/460
monitoring system. This system is a combination of a DuPont Model 400
photometric analyzer and a DuPont Model 460 four point sampling system.
The DuPont analyzer four point sampling system incorporates heated sample
lines, sample probes and a switching network that sequentially samples
from each of the four sample points every 15 minutes.
Oxygen levels were monitored using a Scott Model 150 oxygen
monitor. Sample for this monitor was taken from the DuPont monitor
sampling system so that all four locations could be sampled using one
oxygen instrument without installing additional sample lines.
3.1 DUPONT 400/460 S02 MONITORING SYSTEM
The DuPont S02 monitor is an extractive type instrument where
sample is drawn from the duct to the detector cell by air driven aspirators.
S02 present in the flowing sample cell is detected through photometric
absorption of ultra-violet light in 280-313 manometer range. This U-V
light variation is filtered and detected by both measuring and reference
photo tubes. Measurement of a reference signal compensates for U-V light
variations not due to S02 in the sample. Output from both photo tubes is
amplified and converted to an analog signal proportional to the SOo
concentration in the sample.
Although the DuPont sampling system includes sample dryers,
these dryers do not dry to a constant moisture level. Because these dryers
are located in the analyzer cabinet, which is outside, the sample is dried
to saturation at approximately 20°F above ambient temperature.
Scott Environmental TechnoSosy Inc
-------
C1
a
.1
' n <
i ^
j o
O ~3
S 3
in n>
a>
M
O
1
to
H<
Crt
H
O
Cfl
w
r1
o
n
M
O
2
CO
CANE RUN STATION NO. A SCRUBBING PLANT GAS FLOW
Balanced Draft Fan
Boiler
Electrostatic
Precipitator
Electrostatic
Precipitator
Dampers
© Test Locations
H
(D
.CO
o
o
su
rt
H-
O
3
CO
Balanced Draft Fan
Reheaters Being Installed Here
pi
H
o
M
O
00
Supply to Reactor
1
Quencher
*&r
r
*5-
Scrubber
Module
Return Solution
To Reactor
Supply to Reactor
Return Solution
to Reactor
-------
-10-
SET 1677 01 0678
The S02 system at Louisville Gas & Electric incorporated the
DuPont Model 460 sampling network. This network included heated sample
lines to each of four locations and a switching system that sequentially
samples from each point every 15 minutes. The analyzer automatically
switched ranges from 0-4000 ppm at the inlets to 0-500 ppm at the outlets.
After sampling at each point the DuPont system backpurged compressed air
through that particular sample leg including the sample cell, sample line
and probe. During this backpurge cycle the instrument automatically zeroed
itself using the purge air in the cell as zero gas. Instrument responses
to both sample and zero for each point were recorded on a multi-point strip
chart recorder.
3.2 SCOTT MODEL 150 02 MONITOR
Oxygen levels were monitored using a Scott Model 150 Oxygen
analyzer. The detector in this instrument responds to the paramagnetic
property of oxygen producing an output signal proportional to the oxygen
level in the sample. Instrument responses were recorded on a .strip chart
recorder.
3.3 09/SO, SYSTEM
£~ &.
To facilitate the use of a single oxygen instrument to measure
four duct locations, the sample for the oxygen monitor was taken from the
DuPont monitor sampling system. This arrangement was desirable because the
SOo and 02 analyzers are seeing the same sample simultaneously and a single
analyzer could be used without installation of additional sample lines and
probes. The use of the oxygen monitor in conjunction with the DuPont S02
monitor was invaluable in checking the proper operation of the system as
a whole. For example, small leaks in an individual sampling leg were easily
pinpointed due to the affect on the Q£ level. In certain situations these
small leaks may have gone unnoticed without this configuration.
Scott Environmental "technology Inc.
-------
-11-
SET 1677 01 0678
The combination of Q£ and SO- systems did however present several
problems requiring system design modifications during the program. These
problems were primarily due to the delicate flow system of both analyzers.
The DuPont analyzer employed a balanced vacuum system where a single vacuum
source was used to accomplish several tasks simultaneously. In addition to
drawing sample through one sample line there was a bleed flow of sample
through the remaining three sample lines and flow through all four sample
dryers. System vacuum was held constant by a bleed air regulator that
allowed ambient air to enter the system between the measuring cell and
aspirators in order to maintain a constant vacuum. This balance was easily
upset by drawing sample off for the Q£ analyzer which had to be maintained
at atmospheric pressure.
Figure 2 shows the initial configuration of the SO^/O- system.
This system used a diaphram pump to draw sample from immediately downstream
of the DuPont measuring cell. A portion of this flow was routed through the
02 monitor and the remainder was vented to atmosphere. A solenoid was
installed in the line to the 0~ system to prevent compressed air from
damaging the 0~ instrument during DuPont backpurge cycles.
In addition to many instrument and recorder failures during the
period prior to 11/17/77, it was found that partial or total blockage of a
DuPont system sample leg would result in the Q£ system pulling ambient air
in through the DuPont vacuum regulator thereby producing erroneous readings.
At this point two other problems began to manifest due to decreas-
ing ambient temperatures. These were that the DuPont dryers were not drying
the sample sufficiently for the &2 analyzer and the diurnal temperature
differences were causing Oo instrument drift because it was located outside
with the DuPont monitor.
On November 17 the original Q£ analyzer was replaced with another
Scott Model 150 and the system was modified and moved inside the boiler
house. A schematic diagram of the modified system is shown in Figure 3.
Modifications included taking sample from the DuPont 460 system immediately
Scott Environmental "technology Inc.
-------
J O
O -I
6 3
U) to
O
a
?d
w
M
2!
M
H
M
o
o
M
o
c-
i
o
N3
CO
O
CO
H
§
co
M
H
o
I-1
o
oo
Flowmeter
02 System
Pump
DuPont
Ultra-Violet
Sample Cell
Sample Line
Switching Manifold
Flowmeter
r.
Sample Lines
From 2 Inlets
and 2 Outlets
N3
I
DuPont
Auxiliary
Air Regulator
To Aspirators
-------
SET 1677 01 0678
-13-
Sample Line
Switching Manifold
DuPont
Ultra-Violet
Sample Cell
To Aspirators
Sample Lines
From 2 Inlets
and 2 Outlets
Control Valve
, , _ £ . .
Alternate Configuration
6
<
Solenoid
„ n
Valve
Heated
Sample
Line
Contained in DuPont Cabinet
Located Inside Boiler House
Air Operated
Ball Valves
(Solenoid Actuated)
(^) Valve
Vent
Ive
Refrigerated
Coil Condenser
Vent
Micron
Filter
•Pump
Monitor
QO Automatic
T H20 Purge
V Solenoid
Inc.
FIGURE 3 S02/02 SYSTEM CONFIGURATION AFTER 11/17
-------
-14-
SET 1677 01 0678
after the sample line switching manifold and the installation of a refrig-
erated coil condenser to dry the sample. Heated sample line was used to
deliver sample to the Q£ system inside the boiler house. This modified
system worked well for the remainder of the program. Daily maintenance was
required however to maintain system flow balance. An alternate of this
system where the vented O? sample was returned to the DuPont analyzer was
tried, however, this reduced the flow through the DuPont cell and was there-
fore abandoned.
Scott Environmental "technology Inc.
-------
-15-
SET 1677 01 0678
4.0 OPERATING PROCEDURE
Routine non-maintenance operation of the S02/02 monitoring
system at Louisville Gas & Electric included system calibration and data
collection. The procedures were changed somewhat by the scope of work
change on November 28, 1977. Prior to that date only weekly site visits
were made and calibrations and data collection were done at that time.
After November 28th, calibrations and data collection were done on a
daily basis, Monday through Friday. Additionally, Scott was responsible
for reducing the strip chart data before forwarding it to Monsanto Research
Corporation.
4.1 CALIBRATION
Calibration of the DuPont S02 monitor was done by introducing
both zero and span gases at the sample probe near its entrance into the
duct. Gases were introduced through a calibrate/sample valve located at
the probe. This procedure was used because it is recommended in the
Standards for New Stationary Sources (Federal Register Volume 40, Number 194,
Monday, October 6, 1975) and also because it serves as a check of that
entire system.
This procedure was changed, however, midway through the program
for the following reasons:
9 Since the span/zero gases were introduced in only one location,
this did not serve as a check of the remaining three sample
lines. Calibration at all four locations was impractical.
e There were situations where this procedure of introducing
calibration gas at the sample ports did not show up a system
leak that was present. This leak was isolated through use
of the Oo monitor.
Therefore, during the latter part of the program, calibration gases
were introduced at the analyzer. Leak checks were performed by closing the
valve at the end of each sample line and observing the flowmeter in the
Scott Environmental "technology Inc.
-------
-16-
SET 1677 01 0678
DuPont monitor. Also, the Oo monitor would indicate a significant difference
with even the slightest leak.
The oxygen monitor was calibrated at the same time as the S02
analyzer by introducing nitrogen, 11.5% 02 span gas and ambient air into
the DuPont monitor system and observing the response on the ©2 monitor.
Calibration and leak checks of each sample line were performed
daily which required one man-hour.
4.2 DATA COLLECTION AND REDUCTION
Data collection simply involved removing the strip charts from
both S0« and ©2 analyzers and obtaining the scrubber control room logs
from Louisville Gas & Electric personnel.
For each of four sample locations, an S02 and 02 data point was
recorded four times each hour. Therefore 5376 data points had to be manually
read each week in addition to recording process notes and boiler loads. S0?
data were recorded as percent of full scale and 02 data were recorded as
percent 02.
Data reduction, required after November 28, 1977, was a significant
task requiring approximately 50 man-hours per week.
Scott Environmental "technology Inc.
-------
-17-
SET 1677 01 0678
5.0 DISCUSSION OF RESULTS
The overall operation of the monitoring program at Louisville Gas
& Electric went according to schedule with the exception of chronic mainten-
ance and application problems with the 02 monitor. This section discusses
the routine and non-routine maintenance aspects of both S0« and 0~ monitors.
£ £•
5.1 ROUTINE MAINTENANCE
Aside from the chronic non-routine problems with the ©2 system, the
maintenance of the 02/S02 system at Louisville Gas & Electric was minimal. Approx-
imately 160 manhours for routine maintenance were expended in a 20 week period.
Scott Model 150 02 Monitor
Routine attention to the Q£ system included weekly cleaning of
sample filters and an internal capillary tube in the instrument. The
refrigerated coil condenser installed on November 17th required the addition
of water approximately once per week. These items would require 4-6 man-hours
per week assuming that no additional problems were encountered.
DuPont Model 400/460 SO^ Monitor
The DuPont analyzer at Louisville Gas & Electric required little
maintenance throughout the program. Regular weekly maintenance was limited
to a thorough inspection of the analyzer cabinet components for any signs
of leaks or failure.
Several other maintenance items were routine but not regular.
These included keeping all sample probes clear, cleaning of the monitor cell
and optical windows, and maintenance of the aspirators. In all three of
these cases there would usually be gradual operational indications of the
need for specific maintenance.
Gradual reduction of the flow in the affected sample leg would
indicate a need for cleaning. This gradual reduction, under normal circum-
stances, was enough warning to clean the probe before it was completely
clogged. It is difficult to estimate the time spent on cleaning clogged
probes. During the winter months however probe cleaning became a signifi-
cant task requiring almost daily attention. Maintenance records show that
clogged probes resulted in some lost data.
Scott Environmental "fechnotogy Inc
-------
-18-
SET 1677 01 0678
Accumulated dirt in the sample cell would necessitate compensation
by the automatic zero mechanism in the instrument1^ control unit. When this
fine zero potentiometer would gradually shift significantly from its initial
set points the sample cell was in need of cleaning. This cell required
cleaning only three times during the program.
The need for maintenance of the aspirators was evident by the
gradual loss of excess system vacuum. This could be monitored by a regular
check of the total system vacuum. After the aspirators were cleaned at the
beginning of the program no further maintenance was necessary.
System downtime due to routine component failure can be minimized
by keeping certain spare items in stock. These include:
e Spare probes
o Ultra-violet lamp
e A rebuild kit for the air driven ball valves or a spare unit
e A spare solenoid actuator for the ball valves.
5.2 NON-ROUTINE MAINTENANCE
DuPont S02 ANALYZER
The DuPont analyzer required little non-routine maintenance
during the program. Aside from chronic probe clogging there were only two
instances of component failure during the program. These include the replace-
ment of a ball valve in the sample line switching manifold and the replacement
of a heated sample line which had burned out during the January 24th to
April 4th scrubber outage. Approximately 24 manhours x*ere needed to correct
these non-recurring problems during the 20 week test period.
Scott Oxygen Analyzer
Non-routine maintenance of the oxygen monitoring system was in two
general areas: maintenance of the monitor itself and that associated with
the sampling system and its interface with the DuPont monitor. Table 2
is a maintenance log for the Q£ system showing the major system malfunctions
and their remedies. Aouroximatelv ISO manhours were needed to correct these
problems during the 20 week test period.
Scott Environmental "fechnotogy Inc
-------
-19-
SET 1677 01 0678
TABLE 2 02 MAINTENANCE LOG
Date Problem
8/16 Off scale negative.
No flow to monitor.
8/25 No flow to monitor.
9/16 Source lamp burned out.
9/27 No flow to monitor. Internal check
valve jammed closed. Capillary
plugged. High 02 readings.
10/6 Q£ not responding correctly due to
source alignment problems.
Leaky bellows in pump.
10/13 02 monitor would not calibrate
properly. Found poor connection on
detector.
10/26 Erratic operation of 02 monitor.
11/8
11/16
1/4
1/6
Monitor drift.
02 monitor indicating 70% full
scale at all times.
No flow in 02 system due to pump
failure.
No flow in Oo system due to plugged
valve.
Remedy
Readjust panel controls.
Adjust control valves.
Install larger pump.
Return to lab and replace lamp.
Monitor back in service 9/21.
Cleaned & adjusted check valve
and capillary.
Leak found in DuPont system.
Realign source.
Replace pump.
Realign source. Repair wire.
Realign source. Adjust flow
rates.
Replaced monitor with another
Scott Model 150. Installed
heated line and condenser and
moved monitor inside boiler
house.
Repaired air leak in pump.
Replace pump.
Clean valve.
Scott Environmental "technology Inc.
-------
-20-
SET 1677 01 0678
Problems associated with the monitor usually involved the detector
mechanism and its alignment. The detector employed a light source that
reflected off a mirror onto split photo-cells. Movement of the mirror due
to the paramagnetic property of oxygen in the sample would be sensed by
the photo-cells and a feedback current would be applied to the mirror to
maintain its original position with respect to the photo-cells. This feed-
back current, which was proportional to the Oo in the sample, was then
amplified for output to the recorder. Improper alignment of this detector
would result in erratic and drifting 02 readings. After each realignment
the monitor would work properly for a short period but then the problem
would gradually recur. On November 17th the monitor was replaced with .
another Scott Model 150. This monitor held its alignment for the duration
of the program.
Maintenance problems with the 02 flow system and its interface
with the DuPont S02 system included component failures, flow adjustment
problems and flow blockage problems.
Component failures include the replacement of three pumps. The
corrosive nature of the sample and the need to pull low flow rates against
the vacuum in the DuPont system contributed to these failures.
Blockages in the system were common at several critical points in
the system. These included flow control valves, a small glass capillary
inside the 02 monitor and a check valve also inside the 02 monitor. These
blockages were often due to the formation of solid sulfur compounds rather
than particulate matter.
Flow control also presented problems because while the 02 monitor
required a low flow (approximately 90 cc/min), the bypass flow in the 02
system had to be kept at about 2-3 SCFH to reduce lag time. Bypass flows
above this level would interfere with the proper operation of the DuPont
monitor. Therefore prior to November 28, when Scott personnel remained on
site to attend to these adjustments daily, these critical flow adjustments
would drift resulting in lost 02 data.
Scott Environmental Technology Inc.
-------
-21-
SET 1677 01 0678
6.0 RECOMMENDATIONS
The S02/02 monitoring program at Louisville Gas & Electric Company
provided an opportunity to examine some of the difficulties that can be
encountered in a continuous monitoring program of this nature.
The major difficulties at Louisville Gas & Electric were
associated with monitoring oxygen and maintenance of the DuPont 460 sampling
system which was prone to clogging.
The oxygen monitoring problems were generally a result of taking
sample from a balanced flow system of another analyzer and the use of an
oxygen monitor that did not perform well in a harsh process environment.
As a result more attention should be given to the selection and
installation of the oxygen monitoring system. Selection of a reliable
monitor designed for this type of use is essential. Additionally the
monitor selected should be adaptable for integration into the monitoring
system being used. An integrated system whereby the same sample is being
analyzed for both 0^ and SOo was found to be very useful as an overall
system check.
The DuPont sampling system problems, which were primarily clogged
probes, could be reduced by the installation of effective high pressure
blowbacks at each sample location. Also, for cold weather environments,
attention should be given to heating all areas of the sampling system prone
to freezing.
Manual data reduction for a four point S02/02 system is a major
task. The selection and installation of a microprocessor which logs all
raw data in addition to computing the emission rate in pounds per million
BTU would be very worthwhile.
Scott Environmental "technology Inc.
------- |