Page 313 - Instrumentation Reference Book 3E
P. 313

Temperature measurement considerations 297
                                                      Fractional heating  Where the process fluid flows
                                                      past a probe at high velocity there is, especially in
                                                      the case of gases, a frictional heating effect. The
                                                      magnitude of the effect is not easily evaluated but
                                                      it  is advisable  if  possible to site the  probe  at a
                                                      location where the fluid velocity is low.

             Figure 14.67  Taylor thermospeed separable well   Coizdzictive  cooling  Resistance  thermometers
             system. CourtesyABB Instrument Group.    and  thermistors  depend  for  their  operation  on
                                                      an  electric current  flowing through  them.  This
             can cause a significant measurement  error, caus-   current causes a small heating effect in the sensor.
             ing too high a reading. In the situation where the   When such a sensor is used for liquid temperature
             outer jacket of the vessel is used for cooling the   measurement  the relatively high  specific heat  of
             vessel, for example,  a  cooling  water jacket,  the   most liquids ensures that this heat is removed and
             heat flow will be away from the sensing probe and   the sensor temperature is that of the liquid. How-
             consequently  the  error  will  be  a  low  measure-   ever,  in  gas  measurement  the  amount  of  heat
             ment.  This  conduction  error  is  only  significant   removed is a function of the gas velocity and thus
             where the thermometer pocket is short or where   a variable source of error can arise dependent on
             the pocket is inserted into a gas stream. To mini-   flow rate. In a well designed instrument this error
             mize the error the length of the pocket should be   should be very small but it is a potential source of
             at least three times the length of the sensitive area   error to be borne in mind.
             of the probe.
               The  use  of  a  thermowell  or  pocket  will  also
             slow down the speed of response of an instrument   Cavitation  Liquid  flowing past  a  thermometer
             to  temperature  changes.  A  directly  immersed   probe  at high speed is liable to cause cavitation
             thermometer  probe  will  typically  reach  thermal   at the downstream side of the probe. Apart from
             equilibrium  within  30  to  90  seconds.  However,   any heating effect of the high flow rate the cavita-
             the  same  probe  in  a  thermometer  pocket  may   tion will generate noise and cause vibration of the
             take  several minutes  to reach  equilibrium.  This   probe.  This vibration  is  likely in  due course  to
             delay to the instrument response can be improved   cause  deterioration  or  premature  catastrophic
             in those cases where the pocket is mounted verti-   failure of the probe.
             cally pointed downwards, or in any position where
             the closed end is generally lower than the mouth,   14.7.2.5  Swface temperatwe nieasurement
             by filling it with a heat-transfer liquid. This liquid
             is usually a silicone oil.               Where the temperature of a surface is to be mea-
               An alternative method  for improving the rate   sured this can be done either m7ith  a temperature
             of heat transf'er between the pocket and the bulb is   probe  cemented  or  clamped  to  the  surface  or
             illuskated in Figure 14.67. A very thin corrugated   where a spot measurement is to be made a sensor
             aluminum or bronze sleeve is inserted between the   can be pressed against the surface. In the former
             bulb and pocket on one side. This forces the bulb   arrangement, which is  likely to be  a  permanent
             over  to  the  other  side,  ensuring  metal-to-metal   installation, the surface in the region of the sensor
             contact  on  this  side- while  on  the  other  side,   itself can be protected  from heat loss by  lagging
             the sleeve itself, being made of aluminum which   with thermally insulating material. Provided heat
             has a high theimal  conductivity, provides  a sea-   losses are minimized the measurement  error can
             sonable path for the heat.  In addition,  the bulb   be  kept  small.  Errors  can  be  further  reduced
             should be placed well down the pocket to reduce   where  the  sensor  is  clamped  to  the  surface  by
             the possibility of errors due to heat conducted by   coating the surface and the sensor with heat-con-
             the pocket to the outside with consequent reduc-   ducting grease. This grease is normally a silicone
             tion of the temperature at the bulb.     grease heavily loaded with finely ground alumina.
               The errors associeted with thermal conduction   A grease loaded with beryllium oxide has better
             along the thermometer pocket are of course more   heat  transfer  properties.  However,  since  beryl-
             critical  in  the  case of  gas  temperature  measure-   lium  oxide  is  very  toxic  this  grease  must  be
             ment, as the thermal transfer from gas to therm-   handled with the greatest of care.
             ometer is not nearly as good as it is from liquid.   Where  spot  measurements  are  to  be  made;
                                                      using, for instance, a hand-heid probe. it is dificult
                                                      to get accurate readings. The normal practice is to
             14.7.2.4  Effect of process fluidflow rate   use a probe mounted on a spring so that it can take
             Two  sources  of  error  in  temperature  measure-   up  any  reasonable  angle  to  press  flat  against
             ment are clearly identified.             the surface to be measured. The mass of the probe
   308   309   310   311   312   313   314   315   316   317   318