EPA-AA-EOD-80-13
                         Technical Report
                          December,  1980
                           ASSESSMENT OF
                             TEST CELL
                       HUMIDITY MEASUREMENT
                                AND
                              CONTROL
                                by
                          Sherman D.  Funk
                              NOTICE

Technical reports do not necessarily represent final EPA decisions or
positions.  Their publication or distribution does not constitute any
endorsement of equipment or instrumentation that may have been evalu-
ated.  They are intended to present technical analysis of issues using
data which are currently available.  The purpose in the release of such
reports is to facilitate the exchange of technical information and to
inform the public of technical developments which may form the basis for
improvements in emissions measurement.
                           Engineering Staff
                    Engineering Operations Division
                  Mobile Source Air Pollution Control
                    Environmental Protection Agency •
                          2565 Plymouth Road
                      Ann Arbor, Michigan  48105

-------
                                 ABSTRACT

The  accuracy  of  measurement  and  degree  of  control  of  humidity during
light  duty emissions  and  fuel  economy  tests  have  become  matters  of
increasing concern.  Specific humidity values are used in the calculation
of NOx test results.  It is also  possible  that  in some vehicles humidity
may have an effect on the fuel economy test.

Comparison tests  were made with the  dew  point  method and  the presently
used wet  bulb  psychrometer.   It was  found  that  the wet  bulb  method can
produce significant errors primarily due to  factors  in the  wet bulb tem-
perature measurement.  The dew point  hygrometer  proved  to  be accurate,
reliable, and easy  to calibrate.   It is the  preferred  method  to measure
humidity in the vehicle  test  cells.

A series  of  humidity measurement location comparison  tests  were  made at
three  locations  in  Room  515.  With  proper  seals maintained  around the
overhead doors, the  center of the room proved to be  a suitable  location
to measure humidity.

Tests were made  on the  room's  air handling  system.   A  few improvements
are suggested on the humidity control.

-------
 I.   INTRODUCTION

     This paper  is part  of a  continuum  to earlier  investigations on
     this subject.   It  relates  to  and  answers  questions  that  were
     raised  on the earlier  work.   It  not only  supplements  some  of the
     earlier data  but  also provides some  additional information  in  a
     more accurate manner under  actual  test  conditions.
II.   BACKGROUND

     In late  1979,  automotive  manufacturers  claimed that  the  EPA lab-
     oratory's procedural,  equipment,  and environmental  changes since
     1975 had  a  detrimental effect on fuel  economy measurements.  One
     of  these changes  involved humidity  levels.   In  1975,  specific
     humidity  levels  in  our  test  cells  averaged  about 50  grains  of
     water per pound  of air.   In  1977 these levels were  raised to  75
     grains  to  control  the   NOx   correction   factor  closer  to  1.0.
     Although the manufacturer  claims were not conclusively proved, the
     Administrator ruled  that  we  would lower our humidity levels back
     to the  1975  conditions, (50 grains).  These  claims  also indicated
     there were areas in  our measurement  procedures and  equipment that
     needed  further investigation.   Therefore, we designed a test plan
     that would provide us more information on humidity measurement and
     control.

     This investigation was completed and a report written on April 17,
     1980.  The report included the comparisons of readings of humidity
     measured with three wet bulb-dry  bulb units, one  in front  of each
     dynamometer and one in the center of the room.   These tests  showed
     the following characteristics  and  observations:

     A.  A definite trend of lower humidity  was seen at  Dyno  1  at the
         75 grain level in  the range of a  19 grain  average.   At  the 50
         grain level the average was 9  grains lower.

     B.  Humidity levels  always  increased during highway  fuel  economy
         tests.

     C.  Stability of room  temperature and humidity improved  at  the 50
         grain level.

     D.  Wick contamination and angle  of air flow  across  the wick can
         cause a  2-3  degree F  error on wet  bulb  units.   An additional
         error of  1  degree or  more can  be  realized from thermocouple
         error (J thermocouple).

     E.  Humidity control for the air handler became a source of suspi-
         cion when:   a) outdoor conditions change,  controls  would have
         to be re-adjusted; b)  humidity would increase during high heat
         load, such as the HWFET tests.

     F.  A 1.0°F  change in  wet bulb measurement can result in approxi-
         mately a 2% change in  the  NOx  correction  factor.

-------
      G.   A number of other laboratories are using the more accurate dew
          point meter to measure humidity.
III.  TEST PLAN
      After the  first  report was  completed,  it was  determined  that we
      needed to investigate humidity further in our test cells.  Surveys
      were  made  with  instrument  manufacturer  and automotive  manufac-
      turers  to  gather  information  on  available  instruments.   The
      General   Eastern  Model  1200  APS  Dew  Point Meter  was  selected
      because   of  the  experienced  usage  available,  its  claimed  reli-
      ability,  accuracy  and  delivery schedule.  Three  instruments were
      purchased and set  up on the bench.   A  test  plan was  designed to
      include   check  out  and  evaluation  of  the  instruments,   tests  to
      identify  and characterize  any differences in measurement between
      the wet  bulb psychrometer and  the  dew point  meters,  tests to find
      best cell location for humidity measurement,  further characteriza-
      tion of  the  room air handling control  and  the development  of an
      interactive LCS  computer  programs that  will  calculate  humidity
      from  inputs  of  barometer,  dry bulb,  wet bulb and/or  dew point
      temperatures.

      Evaluation,  calibration  procedure,  and  check  out of  the General
      Eastern   dew  point  meters  was   done   in  a  series  of   steps.
      Information on calibration,  maintenance and reliability  was col-
      lected  from General  Motors   and   General   Eastern  Corp.   Bench
      checking  was done  by using ambient and  ice bottle temperatures to
      check calibration  as  described  in the  instruction   manual.   An
      operational  check  was   devised  by  blending   various levels  of
      humidity  in air  in a sample bag and passing  it  through the instru-
      ment.  Comparisons were made  to  other  instruments in  the lab and
      to a certified  thermometer  accurate to _+ 0.2°F.

      Tests were  made  to  determine  the  difference  in   measurements
      between  the wet  bulb and  the dew  point  method.   Sixteen point to
      point comparisons  were  run.   Since the dew point meter  showed a
      definite  trend  in the negative direction a set of controlled para-
      meters tests was designed and  run.  This was an experimental test
      set-up,  devised and installed in a  small  room with  a  closely con-
      trolled   (humidity  and  temperature) environment.  The  set-up used
      three GE  dew  point units measuring in  series,  one  Sargent-Welch
      wet  bulb-dry  bulb  aspirated  psychrometer,  a  calibrated  mercury
      thermometer,  clean wicks,  temperature  controlled  distilled  water,
      a heated  wire anemometer,  a variable speed blower, and a Honeywell
      temperature recorder.  Comparisons  were made with the  wet bulb to
      the calibrated  thermometer  with and without the  wet sock, with the
      same  velocity  of  air  and  with  the higher  recommended  velocity,
      with ambient temperature for  wick  moisture  and with  water  cooled
      to 60°F or just  above the wet bulb reading.  (Results   are shown in
      Appendix  D and Dl).

      Tests were  made  to find  the best  measurement  location  and  to
      characterize the room  control  system  by  installing  three General
      Eastern Model 1200 APS  dew  point meters  in Room  515, one sensor in
                                  -2-
                                                                      — -•	•TTIW«WT(—• "

-------
     front of the cooling fan  for  Dyno  1,  one in front of Dyno  2,  and
     one in the  center  of the  room.  Also a Sargent-Welch  (wet  bulb)
     unit measured humidity  in the  center of the room.  Eleven certifi-
     cation FTP's  were  measured along  with  six  Highway Fuel  Economy
     Tests (HWFET).

     Additional   room  control  characterization  was  done  by  making  a
     series of various humidity settings in  Room  515 on a day  when no
     vehicle tests were  scheduled.   This was done to determine response
     of change and stability.  A probe  connected  to  the dew  point sen-
     sor was  traversed  across the vertical  inlet duct and  across  all
     room  air  inlets to  determine  humidity mix in  the  air.   Also,
     heated air  from a heat  gun was blown on the room's humidity sensor
     to determine if any effect  on  room  humidity.

     An interactive humidity  calculation was  developed and  implemented
     on LCS that  will provide specific  and  relative  humidity,  partial
     vapor pressure and NOx  correction  factor by inputting,  wet  bulb,
     dry bulb and/or  dew point  temperatures.  This  makes  information
     immediately available for experimental  tests.
IV.  SUMMARY OF RESULTS

     The results of the major  objectives  of  the test plan  can  be sum-
     marized as follows:

     A.   The initial  bench checks  and calibration  procedures  of  all
         three  dew point meters  show that they measure  temperature  at
         ice point  and ambient to an accuracy of _+  0.5°F and that they
         correlate  to  each  other  to  less than 0.5°F difference.

     B.   Four methods  were found to  be acceptable  to check calibration
         of the General  Eastern unit.  They are  the ambient temperature
         check, the ice  point  method, the  blended  sample bag method and
         by  maintaining  an   NBS traceable  unit  in  the  calibration
         department.

     C.   Differences  in  the  wet bulb  psychrometer  and  the  dew  point
         meter  ranged  as  high  as 15  grains but  averaged  5.6 grains with
         the dew point  meter  normally  lower (See  Appendix C  and  D).
         The dew  point   meter method  is   considerably  more  accurate.
         Error  in measurement  of wet bulb-dry bulb can be attributed to
         thermocouple  error,  contamination of wick,  angle  and velocity
         of air flow across wick, and temperature of water.

     D.   As long as there  are no outside  leaks to  the  room, humidity
         levels at  the vehicle  cooling  fans  and at  the  center  of room
         locations  prove to be within  1°F dew  point  (See  Appendix B).
         Specific humidity  levels continued to rise during HWFET's.

     E.   The facility  humidity control system appeared  to  have  a  dead-
         band  that  is extremely  wide.   This  results  in  no response  to
         routine external humidity  changes.   It  was found that for any
                                 -3-

-------
        given humidity setting, the system sensor  controls  on  a rela-
        tive  humidity  rather  a   specific   humidity.    This   causes
        specific humidity to rise during  the  high  heat  load condition
        of the Highway Fuel  Economy  Tests  (HWFET).

    F.  There is no  evidence  of humidity stratification  in the inlet
        ducts.
V.  DISCUSSION

    The EOD study on humidity in the Light Duty  test  cells en which a
    report was written in April  1980  led  to  the  determination to pur-
    chase three  General  Eastern Model  1200  APS  dew  point meters  for
    in-house evaluation.

    These units use the basic principle of dew condensation at a tem-
    perature relative to the specific  amount  of  humidity per pound of
    air.  They use a  platinum RTD  thermometer accurate to ^0.2°C (See
    Appendix A)  to  measure surface  temperature  of a  mirror  on which
    the  dew  collects.    The  mirror  is  thermoelectrically  cooled.
    Cooling is  controlled  by a  photodetector which  senses  amount  of
    light reflectance from the  mirror.  A  light  emitting diode pro-
    vides the  source  of light  that  is directed  on  the  mirror.   The
    units also have  a self cleaning mirror  feature  which works auto-
    matically or manually.

    These units  were  extensively bench checked  upon  receipt  by con-
    necting the  sensors in  series,  removing the  cooler  fuses,   and
    comparing  ambient temperature  readings   to   a certified  mercury
    thermometer.   Units  were also checked at  the  ice  point by flowing
    air  through  a coil  in  a  bottle of  ice  and  through  the  sensors.
    Response  to  humidity  changes  and  stability  was  also   checked.
    Units were  run  at  three operational  levels  of   90,  75,   and  50
    grains  of water  as measured  from  a  blended  sample  bag.   This
    method was developed in-house by  calculating  volume  and weight of
    different  blends,  injecting  measured  amounts of  water  into a
    blended time measured  bag  of air.  After  a  blending  time period,
    sample was drawn  through the DP meter sensors  and dew point tem-
    perature measured.  This value was Lhen used to calculate  specific
    humidity  (grains  of  H2)  per  pound  of   air).   Even  though this
    procedure needs further refinements, all  tests were within accept-
    able tolerances  of advertised specifications.

    After initial check  out of  the  units comparison  tests were made
    with  the  presently  used wet bulb  psychrometers.   Differences  in
    measurement  were  seen  as much  as  15  grains  with the  dew point
    measuring  the  lower.   Under closely  controlled  point to  point
    readings,  the  psychrometer  read 4-5 grains  higher  than  the dew-
    point.   After  extensive   comparisons   confirmed  this   fact  an
    investigation was initiated  to  determine the cause  of these dif-
    ferences.    This   was  done  by  a  controlled   parameters   test  as
    described in the  test  plan  and  results  tabulated   in Appendix D.
                                -4-

-------
The  difference is  mostly  attributed  to  wick  water  temperature
(should be just slightly above wet bulb temperautre ref. NBS Circ.
#512) and air velocity across the wick (should be a minimum of 900
feet per  minute,  ref. NBS  Cinr.  #512  and  Humidity  Measurements,
Instrument Technology).  Wick contamination and  angle of air flow
contribute in lesser degrees.

Once   differences   between   our   standard   measurement  devices
(Sargent-Welch) and the new  dew  point  meters  had been  resolved,  a
series of eleven FTP's  and  six HWFET  comparison  tests  on Dynos  1
and  2  were   run  to  determine  a  suitable  location  to  measure
humidity and  to  observe  room humidity control capabilities.  The
tests were performed  during actual  certification tests.  Humidity
levels in the  cell  were  reasonably  stable during the FTP compari-
sons.  However, after a portion  of  the  tests had been run,  the
humidity  traces  for  Dyno   1 became   erratic  and   showed  lower
readings  than  Dyno  2 and  the  center  of  the room.   Finally   it
became  unacceptable  and  the problem  was  investigated.   It  was
finally traced to a damaged  overhead door and a torn rubber seal.
The seal was repaired.

However, air current  leaks then became apparent at the  Dyno 2 door
but of a less  magnitude.  Even though  both doors appeared to have
been damaged  and  are not 100%  sealed, the leakages  are now only
causing transients of less than 1°F dew  point at the front of the
fan  and  was  not considered  to  be a   problem.   (The  test results
during the door  leaking  problems were not  tabulated  nor included
in this report since  that was an  abnormal condition).   During this
investigation,  a  profile of humidity  mixture  of  inlet  air  was
sampled  traversing  a probe  across  the  inlet ciucts  to check  for
possible stratification.   None was found.   The leaking  door prob-
lem could possibly  explain  the differences  that were seen on Dyno
1  as  documented  in the April  1980  report.   Results  of all tests
(FTP +  HWFET)  showed  close  correlation (less  than  1°F dew point
difference)  from each dyno to the center of the  room  (See Appendix
B).   Due to  the  fact that  on  the   HWFET  comparisons, humidity
traces from the three locations  tracked on top  of  each other and
that  on  some  tests   (high  heat  load)   overall  specific humidity
increased as  much  as  10-15  grains.   Hence,  a table  of estimated
average  values  for  these tests  would  be meaningless  and  v.as not
included.

To  further  investigate this  occurrence,  a  stream of  heated  air
from a heat gun was  directed at  the  room humidity sensor.  Within
a  few  seconds steam  v-ss  being  injected  into the loom  st  a very
high  rate  indicating humidity control relative  to  room tempera-
ture.   Further investigation  into  building   engineerng drawings
confirmed this to be  true.   Room humidity controls sense relative
humidity explaining  why  specific humidity  rises  during high heat
load tests.   Further  review of past work done  on room air handling
systems revealed a memo written in October  li<75  by Doug Berg based
on a  study  by Bene  Engineering  attesting to  this  fact.  It also
agrees with this writer's tecommendation  that we should study our
system to determine if we  could  modify  our humidity sensing and
                             -5-

-------
     control system to control on specific humidity levels rather rela-
     tive.  NOx  factor  is calculated  from  a  specific  humidity value.
     Furthermore if a vehicle's  carburetion system  is  calibrated such
     that it would be sensitive to humidity levels, then it is  specific
     humidity values  with which  we  are  concerned,  not  relative (See
     Conclusions/Recommendations)
VI.  CONCLUSIONS/RECOMMENDATIONS

     A.  The General Eastern Model  1200  APS dew point meter  is  a more
         accurate and  reliable  instrument for measuring  humidity than
         our presently used wet bulb  psychrometers.   It  is recommended
         that the three units  be installed, one in each Light Duty test
         cell mounted  on  the  equipment rack with  outputs  connected an
         analog (0-10 VDC) recorder and  the  LCS  Computer.   It would be
         necessary to  modify  the  computer program to  accept  dew point
         input.   Dew point  temperature   could  then  be  monitored  and
         integrated  over  the  period  of   the  test  and  then  calculated
         directly to specific  humidity and NOx factor.  The L&N thermo-
         couple should also be  input  to  the recorder and  the  LCS for
         measurement of ambient  temperature.

     B.  Differences in measurement of the  dew point  meter and the wet
         bulb units  are  the result of errors in  the wet  bulb  units.
         These differences are caused  by  lower air velocity than recom-
         mended across  the wick,  using  ambient temperature  water for
         wick moisture (should  be  just above  wet  bulb temp), and wick
         contamination.   The   high  degree  of  maintenance  attention
         required  causes  a potential  error  situation.   It  is  recom-
         mended   the   wet  bulb  units   be   phased   out   of  service
         completely.  If  the  method of measuring  and  setting humidity
         is changed  to dew  point  it could be  expected the actual test
         cell  humidity  v/ould  average  4-5 grains  higher  than  is  cur-
         rently seen from  the  wet bulb  control  and measurement.

     C.  Air leakage  from the  soak area  through  the  the  overhead door
         seal  can cause   humidity  stratification  at  the  front  of the
         vehicle.    It  is  recommended that  an  additional sensor  and
         sampling kit  be  purchased  for a routine  measurement with the
         sensor mounted on the  cooling fan as part of the Repca diag-
         nostics.   This will minimize  the  possibility  of  any  long term
         problem existing  undetected.

     D.  Relative  humidity  is  the value reported  on emissions  test
         results.    Test   cell  humidity  is  controlled  as   relative  to
         ambient temperature.   Given specific humidity (grains of water
         per pound of  air) is  an engineering unit  value  used to calcu-
         late  NOx factor.  It  should be  the  value  reported.   Also,
         since some  test  vehicles  carburetion  may  be  more  sensitive to
         specific humidity, it  is  recommended a study be  initiated to
         determine the feasibility  of changing the air handling systems
         in the test cell to control on  specific humidity levels.  This
         would prevent the system from adding moisture where  the inter-
         nal heat load is high, such  as  the fuel economy tests.  This
                                 -6-

-------
           recommendation agrees with  the  Bene Engineering  study report
           and Doug Berg memo of October 1975.  Also  it should be inves-
           tigated to determine  if after changing  the  control to specific
           humidity that the deadband width settings could be adjusted to
           be more realistic and  applicable.

       E.  The use  of  the  General  Eastern  dew  point  meter  method  for
           measuring humidity can  provide  reliable  and accurate  data if
           certain procedures of calibration  checks and preventive main-
           tenance  are  followed.   It  is   recommended   the  following
           procedures be used.

           1.  Mirrors in the sensors  should  be cleaned as  prescribed in
               the instruction manual once  a month.

           2.  The  thermometer   should  be  checked  every   90  days  by
               removing the  cooler fuse to allow  ambient  air temperature
               to be compared to a certified mercury thermometer accurate
               to +0.2°F.

           3.  Condensation  and  operation  checks  should be  made  at dif-
               ferent levels (90, 75 and 50  grains,  .   .  .)  by using the
               blended  bag  method  developed  in-house.  This  procedure
               needs to be documented.

           4.  Two  more  units  should  be  purchased  for  accuracy trace-
               ability and  maintenance back up.   One unit should  be sent
               to NBS for certification and maintained in  the Electronics
               Shop as  a standard.  The other  unit would  be used  as  a
               spare or for  additional  studies.
 VII.  APPENDICES

       A.  General Eastern Model  1200  APS  Dew  Point Hygrometer Information

       B.  Humidity Measurement Location Comparisons

       C.  Wet  Bulb-Dry  Bulb vs.  Dew Point  Meter  Humidity Measurement
           Comparisons

       D.  Controlled Parameters  Test  WB-DB vs. Dew Point Method

           Dl  Wick H20 Temp and  Air Velocity  Effect  on WB vs DP Method

       E.  Room 515 Humidity Sensor and Control

       F.  Dew Point/Wet Bulb Temp  vs. Specific Humidity


VIII.  REFERENCES

       1.  Reports and Critique  on In-house Air  Handling Systems.  April
           - Ocotber 1975, Bene Engineering and Doug  Berg.

       2.  Humidity   Measurements,    Instrumentation  Technology,   P.R.
           Weiderhold.

-------
 3.   EPA Memo,  Configuration Standardization  of the  Wet  and  Dry
     Bulb Psychrometers.   Carl Ryan, April  18,  1980.

 4.   EPA Memo,  Change  in  Procedures  1088,713 and  711,  L. Hormes,
     March 4.  1980.

 5.   EPA Memo, Monitoring Humidity control, C.D.  Paulsell, March 6,
     1980.

 6.   EPA Memo,  Change  of Test  Cell Humidity,  C.D.  Paulsell,  March
     4,  1980.

 7.   Humidity Calculation  and  Report  Program Documentation,  Dave
     Holland & Mary  Saelzler, May  13, 1980.

 8.   NBS Circular #512,  Methods  of Measuring  Humidity  and Testing
     Hygrometers,  Arnold  Wexler and W.G. Brombacher, Sept. 28, 1951.

 9.   EPA Memo, Comparison of Humidity Correction  Factor Procedures,
     E.P. Zellin,  May 31,  1973.

10.   GM Memo, Procedures  for Calculating Test  Area Humidity, David
     Horchler, May 31,  1972.

11.   Project Subtask Summary, Dew Point Hygrometer Evaluation, S.D.
     Funk,  Sept. 1980.

12.   NBS Paper, Vapor  Pressure  equation for Water  in  the Range of
     0-100°C,  Arnold Wexler and Louis Greenspan,  Feb. 19,  1971.

13.   EPA Paper, A Procedure  for  Calculating  Humidity,  Eric Zellin,
     July 1975.

14.   Discussion of  Paragraph 85.1320-1 of  Proposed  Rule  Making,
     Eric Zellin,  June  1975.

15.   EPA Paper, Test Cell Humidity Investigation  Report,  Sherman D.
     Funk,  April 17, 1980.

16.   Technical Literature on Humidity Measurements, General Eastern
     Corp., John  Harding and P.R.  Widerhold, June  1978  and  June
     1979.

17.   EPA Memo,  Progress  on humidity Investigation, April  4,  1980,
     S.D. Funk.
                             -8-

-------
                                      SECTS C
                    PRINCIPLE  OF  OPERATION
                                                                           APPENDIX  A

                                                                           Page 1  of 3
2.0   PRINCIPLE OF OPERATION

The Series 1200 Dew Point Hygrometers ore thermo-
elcctrically-cooled, optically-detected, automatically-
controlled, condensation (or "dew point") hygrometers.
The principle of  operation  may  be seen  from  the
accompanying figure.
               O'llCAl OAtAMCE
 The condensate detection mirror is illuminated with a
 high intensity, solid state, light-emitting-diode (LED). A
 photodetector is configured so as to monitor the specular
 (direct) component of the light from the mirror. A separate
 LED and  photodetector  combination are used to
 compensate for any thermally-induced changes in the
 optical components. T.he phoiodetectors are arranged
•in an electrical. pridge'circuit-such.that:the-speculc'r:
- detector is fully illuminated when the mirror is clear of
 dow, and sees reduced light as dew forms on the mirror,
 due to scattering losses.

 An optical offset is designed into the bridge, such that a
 large bridge output current is developed v/henever the
 mirror is in the "dry" condition. The bridge output is
 amplified and used to control the direct current to the
 thermoelectric cooler, causing the mirror to cool toward
 the dew point. As dew begins to form on the mirror, the
 optical bridge isdriven toward its balance point, causing a
 reduction in the specular light, causing  the bridge
 output to decrease, and the cooling current to reduce.
 A rate feedback loop within the amplifier is employed to
 insure critic?.! response, and the system quickly stabilises
 at  a condition wherein a thin dew or frost  layer is
 maintained on tho mirror surface, i.e., the dew, or frost,
 point. A precision thermometer element is embedded
 within the mirror which monitors this dew poin! temper-
 ature directly.

 Ttie condensation hygrometer is a fundamental method
 for measuring  water vapor, affording a degree  of
 accuracy not available in other methods. Additionally.
 the rcpeatabilty of the instrument can be  checked at
 any lime by opening the control loop and allowing the
 mirror to heat and the dew to evaporate, and  then
 reclosing the control loop. Long term accuracies in the
 order of ±0.1'C arc common with  the condensation
hygrometer, making it suitable as a laboratory reference
instrument.

AUTO-REFLECTANCE/PACEFT FEATURE*

The 1200 "S" Series instruments are equipped with
General Eastern's Programmable Automatic Contami-
nant Error Reduction circuitry (PACER") which permits
the instrument to check its own performance and make
any necessary optical adjustments on a periodic basis
(once per 2. G, 12. or 24 hours, or on command). This
circuit automatically places the instrument  into tho
"optical adjust" mode, then cools trie mirror for 30
seconds causing soluble contaminants to dissolve into
the excess water. This circuit then automatically causes
the mirror to heat to a dry condition and, as  the dew
layer evaporates, any remaining soluble contaminants
collect  at isolated sites, leaving a mirror surface with
substantial  surface  area  available for  undisturbed
growth  of condensate. When the mirror is heated to the
dry condition the PACER"  circuit then automatically
adjusts  the offset current to its correct value, regardless
of mirror contaminants. While this function is being
performed, the output sig nal is held at a constant value,
equal to the last  dew/frost point value, allowing full
compatibility with process control loop equipment. The
autoreflectance circuit then returns the instrument to
 normal  operation. Logic signals are provided at the rear
. of the instrument to identify "normal" and/lautoreflec-
 tahce adjust" conditions.   •'..'••' ..:,...••/';"•' :.'•.  •. • •
 PREVAILING . ^.l
 Dirt PCINT
    "I
    s\
                                        OPftCAl
                                        SYSTEM
                                        AU10MArtCM.lv
                                        *o*jsis ron
                                        Connect
                                        AtfltCIANCC
                                        I * SFC |
MI n HQO
ft£TURKt TO
ucw roiMT
               TYPICAL PACER1" CYCLE
            (Timci c;:n be programmed by user)
        Tho 1200 S?rfes Of Gandrjt Eastern Corporation Instruments
        uMl*>ioa fundjoieoui meaiufirvj lochnique. Whan proved
        with pl^Jlnun* ii>si3i>>nce th«rmoti*!or mirror tomnor^ture
        scnto^* wiih NOS if-Cf,.Me r.o^M.cier.ts, |?00 Seitc»
        in*1furr.cnu AO su.t.iMc tew melrolcgtcAl M jndAtOs.
        Ct'/'/•<:«(rJ onre.:fa!>iHtY6rcp(Qvid<*(t v-»;/i «// I200s6fi9*
        seniors Mcorpo/of/nj p:»;inum RTD n//ro/ thtttnomttetf.
   Patents applied for.
                                               2-1

-------
                                                                             APPENDIX A
                                                                             Page  2 of 3
2.1   SUMMARY OF SPECIFICATIONS

2.1.1  RANGE

The basic dew point range of the sensor Is dictated by
the heal pumping capacity of the thermoelectrically-
cooled mirror, and is typically from M-80F to -40F. That
is, with the sensor installed in a typical environment of
+POF, daw points can be measured from the saturation
temperature- (+80F), down to -40F. Below +32F, of
course, the sensor actually measures the frost point
temperature, since dew cannot exist below -f 32F on a
continuous basis. If the  ambient  temperature of  the
sensor is low;:r than 80F, then the lowest measurable
dew point ia lowered. Conversely, if the sensor is
installed at a  higher-than-ambient  temperature,  the
high end of the range Is increased, and the lowest
measurable dew point is raised. (See Range Graph)
      »200
      + 100
 Ul

 $
 U)
      -100
         -100         0         +100

               DE.W POINT TEMPERATURE.
                                           +200
 Each sensor is equipped with an optional water cooling
 jacket which, when supplied with water at approximately
 0.5 gpm, will permit a 0.67CF improvement in the lowest
 frost point attainable, for each 1 °F reduction in temper-.
 ature of the senspr'prpvided by the coolant, For example,",
 if 55'"F cooling water' Vs  available, the sensor ''can'
 measure frost points from +55F down to -56F.

 2.1.2  HIGH DEW POINTS

 The sensor may be operated at dew points up to
 +212:F. as long as the sensor body is kept above the
 highest dew point anticipated to prevent condensation
 on the walls of the sensor. Usually, the gas (or oven,
 kiln, dryer, etc.) has sufficient heat capacity to raise the
 temperature of the sensor body, if properly located or if
 equipped with a simple insulative jacket.

 2.1.3  ACCURACY

 The accuracy of the Series 1200 dew  point sensor is
 determined almost entirely by the accuracy of the
 condensation mirror temperature sensor.

 The mirror is provided with a precision platinum resis-
 tance, thermometer with NBS-traceable coefficients.
 The accuracy is typically, ±0.40'F, as shown in the
 graph.

+ 1.0
u.
o
^_^
i^foyte?-: ;.' >sx.v..w..i..'f ...^.-•...-., ........
^^^!yS>^^:Vv^:^-r-"--:--:^
, ' 1
00 0 +100 +200
DEW POINT. OF
2.1.4  DEW POINT SENSITIVITY

The fundamental sensitivity of all Series 1200. dew
point sensors is ±0.05°F.

2.1.5  DEW POINT RESPONSE

The response of all Series 1200 dew point sensors is a
function of the cooling and heating rate of the mirror
(typically, 3"F/sec), the sample flow rate, and the abso-
lute value of the actual dew  or frost  point being
measured. At dew points above 32F, the response is
almost entirely a function of the mirror cooling rate and
gas flow rate. Below 32F, the response is slowed by the
increasingly reduced availability of water vapor in the
sample as the dew point is lowered, and by the out-
gassing characteristics of sampling lines.  At -40F
frost  point,  the response  is  several seconds. At
—60F, several  minutes  are  required  for a 63%
response to a change, and the lag is-a function of the
outgassing characteristics of the sampling lines, sample
flow,  or availability  of water molecules and  crystal
growth rate.

2.1.6 AMBIENT TEMPERATURE LIMITS

The Series 1200 sensor will operate over the ambient
temperature extrern'es-"of*l-212F to -75F.The electronic
control module is suitable for operation between +120F
to -H32F.

2.1.7 CIRCUIT DESCRIPTION

The 1200 "S-Series" instruments all have essentially
identical circuitry. The only fundamental  difference
between systems is the omission of the digital readout
meter in those systems packaged in NEMA enclosures.

The System consists of several separate subsystems:
(see Functional Block Diagram)

  1. The mirror temperature dew point control loop is
     located on the 1201 CAS printed circuit card. This
     circuit amplifies the signals from the photot ransis-'
     tor detectors in the sensor to a level suitable to
     drive the thermoelectric cooler power amplifier
     transistor. The  final power transistor is mounted
     directly  to the  instrument chassis for heat sink
     purposes.
                                                                                                2-2

-------
                                                                         APPENDIX  A

                                                                         Page  3 of 3
                                Bopitnlo.d wlih porml»!on f'om AWS |R«f. A)
                                  1000
                                        -70
                                                      Frost poinl.'C                      Dew poinl.'C
                                                 -60'  -50   -40  -30   -HO_-IQ__0   10 20 30 40 50
                                •g lOOr-
Figure 3.  Uncertainties In typical
humidity sansors. under laboratory
conditions,  are illustrated  here.
The gravimetric train, an extremely
accurate sensor which weighs ac-
tual moisture content  In known air
volumes. Is used by  NBS to cali-
brate other sensors.  In  industrial
installations. - few  devices  attain
the accuracies  Indicated.
                                                                Typicot dec. ..
                                                                RUCP2D-C
                                                                (Oumno'C, Pope)
                                                                                    Kolled. (reeled
                                                                                    hoif@>25°C
                                                                              LSO». soil ll.iCI)
                                                                                de« poinl(u>25*C
                                                                                     1	
                                 •":',. 0.001-
:.'   O.I .•:...'. . .. .1,
 Mixing'ratio, g/kg
10
100
+200
 + 150 h-
 -too
    -100     -50       0      +50     +100
                        Ambient lemperoIufe/F
                                              + 150
                                                       — 2000 PPM
                                                         -200PPM


                                                            20 PP.M


                                                           ]?. PPM
                                                         + 200
          Figure 2. Useful operating ranges
          of various humidity sensors. Psy-
          chromotors. percent RH. nnd sat-
          urated salt dew point sensors op-
          orate in the mid-ranrjos of temper-
          ature and humidity. Condensation
          hygrometers aro  useful  ovar  a
          largo  temperature span  and to
          humidities as low  as one purcont.
          Electrolytic  hygrometers operate
          at  low clow points over the entire
          range of humidity.

-------
                        APPENDIX B
                    Humidity Measurement
                 Location Comparisons & WB
Date
10-9-80
10-10-80
10-15-80
10-16-80
10-17-80
10-30-80
10-31-80
11-20-80
11-20-80
11-21-80
11-21-80
FTP
Test Pair
Numbers
80-6369
6370
CENTER
6863
6864
CENTER
6388
6375
CENTER
6371
6391
CENTER
6353
6429
CENTER
6586
6621
CENTER
6652
6576
CENTER
6820
6822
CENTER
6816
6823
CENTER
6869
6865
CENTER
6825
6833
CENTER
Barometer
28.86
28.86
29.20
29.05
28.70
29.25
29.98
29.23
29.13
29.14
29.18
Dew Point
Temp.
44°
44°
44°
44°
44°
44°
51°
51°
51°
47°
47°
47°
48°
48°
48°
44°
44°
44°
48°
48°
48°
46°
46°
46°
47°
47°
47°
43°
43°
43°
45°
45°
45°
Dew Point
Spec. Hum
44. OCR
44.0
56.8
49.1
51.6
43.4
49.4
46.9
48.9
41.9
45.2
Wet Bulb
Temp.
59. 5F
59.8
59.0
59.2
59.3
57.5
59.5
59.5
60.3
58.5
58.5
Dry Bulb
Temp.
75. OF
75.5
72.0
76.0
75.5
75.0
76.0
76.5
77.3
77.0
76.5
WB Spec.
Hum
54.4GR
54.1
54.6
49.2
51.4
43.0
48.1
49.3
52.0
44.3
45.0
Hum
Diff
-10.4
-10.1
2.2
-.1
.2
.1
1.3
-2.4
-3.1
-2.4
.2
                                                              AVG.    -2.2
NOTE 1:    WB-DB readings  are results  of operator eye averaging resulting in
           a less accurate  comparison  of WB-DB vs  dew point meter  than by
           point to  point  as referred  to in the  text  report.   Dew Point  was
           most always  lower.


     2:    1st Number of Pair is DOOl    2nd Number  of Pair  is  D002.   CENTER
           is center of room.

-------
              APPENDIX  C
Wet Bulb-Dry Bulb vs. Dew Point Meter
   Humidity  Measurement Comparisons
TEST #
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
BAROMETER
IN HG
28.81
29.00
29.00
29.00
29.06
28.21
29.00
29.00
28.81
29.18
28.88
29.15
29.10
28.94
29.05
29.20
WET BULB
TEMP. °F
60.5
58.0
58.0
58.0
58.5
61.2
58.2
58.0
60.0
57.4
59.0
61.0
58.9
58.3
63.0
60.5
DRY BULB
TEMP. °F
76.0
73.0
72.5
72.0
76.0
76.0
72.2
74.5
75.0
74.0
75.0
76.0
76.0
74.8
76.0
76.7
WET BULB
SPEC. HUM.
55.7
49.1
49.9
50.8
46.1
60.8
51.3
46.6
55.1
44.4
50.5
57.1
47.8
47.6
66.8
53.5
DEW POINT
TEMP. °F
45.8
45.3
44.9
45.6
44.5
45.8
45.0
43.9
45.3
42.6
45.3
45.3
43.6
44.3
48.0
46.9
DEW POINT
SPEC. HUM.
47.3
46.1
45.4
46.6
44.6
48.3
45.5
43.6
46.3
41.2
46.3
45.8
43.0
44.4
51.0
48.6
DIFF. DP HUM
-WB HUM.
- 8.5 GR
- 3.0
- 4.5
- 4.2
- 1.5
-12.5
- 5.8
- 3.0
- 8.8
- 3.2
- 4.2
-11.3
- 4.8
- 3.2
-15.8
- 4.9
DIFF NOx
FACTOR
-.0323
-.0113
-.0169
-.0155
-.0058
-.0490
-.0215
-.0108
-.0332
-.0115
-.0157
-.0439
-.0175
-.0115
-.0643
-.0132
                                         Ave,
-6.1   Ave.
-.0233
      NOTE:   iThis  table  lists  a  portion  of  over  50  comparisons.
              Maximum   and   minimum  differences  are   included.   Most
              differences were 3-4 grains, dew point being lower.

              ^Comparisons  were  point  to point  with  no  vehicles being
              tested.

-------
                                                           APPENDIX D
                                                   Controlled Parameters Test
                                                    WB-DB vs Dew  Point Meter

Test     Wet Bulb     WB Wick       Air Vel.     Bar    Dry Bulb   Wet Bulb     Wet Bulb    Dew point   Dew point   Diff Dp Hum.
  #    Thermometer   Water Temp.  Across Wick  In. HG   Temp."  F   Temp."  F    Spec. Hum.    Temp.0  F   Spec.Hum.   -WB Hum.

01
02
03
04
05
06
07
08
09
10
J
THERMOCOUPLE 75.0
CERT
MERC +0.2F 75.0
CERT
MERC +0.2F 75.0
CERT
MERC +Q.2F 75.0
CERT
MERC +0.2F 65.0
CERT
MERC +0.2F 60.0
J
THERMOCOUPLE 60.0
CERT
MERC + 0.2F 75.0
CERT
MERC +0.2F 60.0
CERT
MERC +0.2F 65.0

650. FPM 29.00 76.0 60.5 55.2 GR 46.9 49.2 GR -8.3 GR
650. FPM 29.00 72.2 58.4 52.2 44.9 45.4 -6.8
850. FPM 29.00 75.0 58.4 47.5 44.9 45.4 -2.1
950. FPM 29.00 75.0 58.2 46.7 45.0 45.6 -1.1
950. FPM 29.00 75.3 58.0 45.3 44.7 45.0 -0.3
950. FPM 29.00 75.6 58.0 44.8 44.6 44.8 0.0
950. FPM 29.00 74.8 58.0 46.1 44.3 44.3 -1.8
750. FPM 29.00 75.5 58.2 45.8 43.9 43.6 -2.2
750. FPM 29.00 74.5 58.0 46.6 43.9 43.6 -3.0
700. FPM 29.00 73.00 58.0 49.1 44.3 44.3 -4.8

-------
APPENDIX Dl
                    WICK H20 TEMP 6. AIR VELOCITY EFFECT
                            ON WB VS. DP METHOD
                                AIR VELOCITY
                                feet per tnin
                                                                    1-15-81
     WB ^- 58°F
               60
WET BULB
H20 TEMP°F     65
               75
650
700
750
850
                                           Decreasing
950
-3.0(M)
-4.8(M)
-6.8(M)
-8.3(J)
0
-1
-0
-2.2(M) -2. 1(M) 1
.000
.8(J)
.3(M)
.KM)
                                                       
-------
:-!-!;; iT:rii-ifi;i;i-:fri!

1 ...
*'"""*"

, : : '
.._..
* . ' '
• - !
1 • '
	 J
! • • 1 '
:.:i. ::!
: • •: • ! ' | •
_ - 	 1 .... J -.
•:•:;.-::!


-._.


" " !

-------
\   v

-------