During the startup of the Orange County (Calif.) Sanitation District’s new trickling filter--–solids contact process, a set of surprising test results revealed a shocking fact about some essential lab equipment.
The process has two, parallel mixed liquor channels. Each channel feeds three secondary clarifiers. The in situ total suspended solids meters for each channel showed significant discrepancies, which caused concern about unequal flow and solids distribution.
Because of limited analytical capability in the field, a comparative 30-minute sludge-settling test (SSV30) was performed using settlometers on each of the two channels. The hope was that this test would provide some indication of whether the solids concentrations in the two channels were as different as the in situ meters indicated.
The SSV30 results for the two channels clearly showed a discrepancy (though not as much as the in situ meters indicated). One channel had an SSV30 value of 110 mL/L, and the other channel had an SSV30 reading of 70 mL/L. (The photo shown was taken at a later date and does not indicate these values.) Visually, however, there seemed to be about the same amount of sludge in each settlometer.
Upon closer examination, the scales of the two settlometers were observed to be different. While the 1000-mL (2-L volume) marks were reasonably close to one another, the lowest marks (50 mL) on each settlometer were different distances from the bottom of the vessel (see photo below).
The analysts conducted a volumetric test of each settlometer using graduated cylinders and concluded that one settlometer had a small error at the 1000-mL (or 2-L volume) mark, but a significant error at the low end of the scale. At the 100-mL (or 200-mL volume) mark, the error was about 30 to 40 mL.
The scale on the inaccurate settlometer seems to be compressed. Both settlometers were manufactured by the same company. The only other difference between the two vessels is that the accurate settlometer has a bottom stamp that reads “Made in USA.” The inaccurate settlometer does not have a stamp indicating where it was manufactured. (The original wrapping was found later and indicates it was made in another country.)
This discrepancy would change the shape of the settling curve slightly but have a potentially significant effect on the SSV30 and sludge volume index (SVI) measurements. If each of the settlometers was used randomly over time, it might suggest process variability that is not really occurring.
If this were an activated sludge process with SSV30 values of 200 to 300 mL, the problems would not be as significant, because the error decreases with increasing volume.
For a trickling filter solids–contact process, however, the error is significant, since the process SSV30s tend to operate at the lower end of the settlometer scale. What is the impact? If the mixed liquor suspended solids concentration is 1500 mg/L, the SVI would be 47 mL/g in one case (inaccurate) and 73 mL/g in the other (accurate).
Lower SVIs are expected with trickling filter--–solids contact (100 mL/g or less) but shouldn’t be less than 50 mL/g. Consequently, what appears to be a manufacturing error might have led to a poor process decision.
Operators and lab analysts should have absolute confidence in the accuracy of glassware. But to quote an earlier edition of Standard Methods, “A good analyst tempers confidence with doubt.”Lesson learned.
The good news is that the flow and solids distribution between the two mixed liquor channels appears closer than the in situ total suspended solids meters suggest, and the SVI seems to be within the acceptable 50 to 100 mL/g range.
Woodie Mark Muirhead
is an operations specialist in the Honolulu office of Brown and Caldwell (Walnut Creek, Calif.).
is vice president in the Phoenix office of Brown and Caldwell.
is a senior engineer at the Orange County Sanitation District (Fountain Valley, Calif.).
©2011 Water Environment Federation. All rights reserved.