The aim of a percolation test is to evaluate the long-term constant rate at which a test location can be expected to accept continuous wetting. It is not readily inferred from characteristics seen at the surface, for example slowness of a puddle to dissipate. A hole must be made to the depth range where percolation is to take place and elaborate pains must be taken to pre-soak the hole and maintain saturation prior to testing. Testing must be repeated and a collection of data is collected allowing judgement of a terminal percolation rate. Several such measured values from across the percolation area are desired, and an average rate is assigned.
Many difficulties accompany percolation testing and data are often thrown off by error-causing phenomena. These include:
clay layers that can be inadvertently smeared over other strata during the drilling procedure (and can make a hole appear to percolate too slowly),
sidewall pieces falling into the hole (likewise making a hole per too slow), and
sub-surface anomalies such as gopher-holes (which can make a hole seem too).
A field professional should consider reaming holes with scratching devices, lining the holes, and overnight pre-soaks with reservoirs of water for continuous feed, and placement of spare holes to replace those compromised by testing difficulties.