Chapter  7—AUXILIARY  MACHINERY gage glass will also result in false level indications in the gage glass. Once the distilling plant is in operation, the feeding must be maintained at a steady rate. Sud- den rising of the water levels or too high a water level will cause carryover of small particles of brine with the vapor (priming). The level of water in the shell must be carried at the highest level that can be held and still prevent the carrying over of saltwater  particles  with  the  freshwater  vapor, otherwise scale will form rapidly on exposed tube surfaces. The pressure differential between the first and second effects permits the second-effect feed to be discharged into the second-effect shell. A par- tial or total loss of pressure differential indicates that air leaks have occurred between the first and second-effect shells in the two-effect distilling plants. Large air leaks between the first effect and second effects can be readily detected because the vacuum gage for the first effect will read approx- imately the same as the vacuum gage for the sec- ond effect. Large air leaks of this type will disrupt the operation of the plant and must be located and  repaired  before  the  plant  will  operate properly. Improper  Venting  of  Evaporator Tube  Nests Improper venting of the evaporator tube nests causes either an accumulation of air in the tubes, with a loss of capacity, or an excessive loss of tube nest steam to the distilling condenser, with loss of  economy.  Troubles  of  this  type  usually  result from   improper   operation   rather   than   from material failures. Scale Deposits on Evaporator Tubes Until 1958, scale deposits on evaporator tubes had  been  one  of  the  more  serious  causes  of operating  difficulties.  In  1958,  a  new  compound was  authorized  for  treatment  of  evaporator feedwater.  The  new  compound  PD-8  evaporator treatment is far superior to the cornstarch-boiler compound formerly used. For details on PD-8 and  its  use,  refer  to  the  applicable  chapter  in Engineman  3  &  2,  NAVEDTRA  10541  (current edition). Last-Effect  Shell  Vacuum Most   manuafacturers’   technical   manuals indicate   that   a   vacuum   of   approximately 26 inches of mercury should be obtained in the last-effect  shell  when  the  temperature  of  the seawater is 85 °F, and that the vacuum should be higher when the seawater is colder. Failure to ob- tain a vacuum of 26 inches of mercury, or more, can generally be traced to one of the following factors: air leaks, improper operation of air ejec- tors, insufficient flow of seawater, and ineffec- tive use of heat transfer surface in the distilling condenser. Testing for Air Leaks The  importance  of  eliminating  air  leaks  can- not  be  overemphasized.  Many  distilling  plant troubles are direct results of air leaks. Air leaks in the shells of distilling plants cause a loss of vacuum and capacity. Extreme care must be taken in making up joints and in keeping them tight. Joints should be periodically tested under pressure for leaks. There are several methods by which tests can be   made   for   air   leaks   in   tube   nests,   heat exchangers, shells, and the piping systems of the distilling plant. When the plant is in operation, a candleflame can be used to test all joints and parts under vacuum. With the plant secured, air pressure tests or a soapsuds test can be used on the various component parts of the distilling plant. The manufacturer’s technical manual describes how the various parts of the plant can be isolated and placed under air pressure. Air   leakage   may   also   be   detected   by hydrostatically testing the various parts of the plant. When performing air tests or hydrostatic tests, precautions should be taken not to exceed the maximum limit of the test pressures specified by  the  manufacturer. Testing  for  Saltwater  Leaks If a leak is detected in a heat exchanger, the defective tube(s) should be located by means of 7-25