REMOVING THE LIMITS ON DEEP-BED FILTRATION
THROUGH NEW BACKWASHING TECHNIQUES
John E. Gay, Group Manager
Hydromation Company
Livonia, Michigan 48150
INTRODUCTION
The state of the art of deep bed filtration, which for many years advanced very
little, has now sprinted ahead, and as a result made deep bed performance available in
many areas of application which had to settle for less effective or far more expensive
methods of filtering.
The problem  of preventing filter degradation as a result of dirt accumulation in the
filter's bed is the area which has yielded to a new approach.   Sharply reducing this
problem means the deep bed filter can now deliver its full effectiveness during 95% or
more, of the hours it is in service because filtered-out dirt can now be disposed of
dependably, quickly and without great cost or effort.
Deep Bed filters embodying the design advances have been placed in service, and
have delivered on the promise in every instance.   Thus the advance has turned from
theory to reality and demonstrated in the process that it can be applied widely with
excellent results.
That is the thrust of this paper.   But, before details are presented, it will be advantageous to first review the filtering situation to see where this development fits in.
SYNOPTIC LOOK AT FILTRATION REVEALS OPPORTUNITY
FOR DEEP BEDS
Look first at the range of filtration needs expressed in terms of desired cleanliness.
The chart in Figure 1 suggests which filter types are capable of practical filtration over
what range of particle-size limits.   Note, for example, that the strainer is adequate to
remove particulate of 1000 ix down to perhaps 100 ix.   The pressure plate is usually
effective if particle size is between 350 ix and 20 /i, or even finer if pre-coat and/or
body feed is used (shaded end of bar).   The vacuum filter has a similar range, and both
the deep bed and pre-coat filter types demonstrate effectiveness between 200 ix and
1 ix.   Filtration below 0.5 microns is the province of reverse osmosis.
Now by adding another dimension to the chart (Figure 2) we can see how each of
these filter types ranks in regard to flow capability.   Note that strainers may flow 100
gallons-per-minute per square foot, if 1000 ix straining will do, but filtration to remove
everything above 5 microns should, up to now, be limited to less than six gallons-per-
minute per square foot of filter area.
Go further and add yet another dimension to this chart, suggesting the cost-per-gallon
comparison between various filter types (Figure 3).   Use it particularly to spotlight the
comparative differences in cost-per-volume unit as you move from coarser to finer
filtration.   Our chart considers both capital and operating cost, and it expresses the
relationship of one filter type to another, rather than attempting to place a dollars-per-
gallon value on each type.
776