Page 262 - Fundamentals of Water Treatment Unit Processes : Physical, Chemical, and Biological

Coagulation 217

(a) (b)

FIGURE 9.20 Jar test apparatus manufactured by Phipps & Bird, Inc. (a) Jar test apparatus with water bath; used for research, c. 1970 and
(b) programmable jar tester, Model PB-900e, c. 2000. (Courtesy of Phipps & Bird.)

Typically, the jar test consists of filling the six jars with
1000 mL each of the sample water. Different dosages of
time, with the modern version of rapid filtration tech-
coagulant are added to the respective jars to span the expected
nology being put into practice only about a decade
optimum dosage range. The coagulants should be added sim-
earlier, the need for initial mixing of alum with water
ultaneously after the rapid-mix is started. The sequence nor-
was recognized as important, but the idea of forming
mally is 3 min rapid mix at 100 rpm and about 15–20 min
flocs with velocity gradients was contrary to the con-
slow mix to simulate flocculation followed by 30 min settling.
ventional wisdom; rather, quiescent settling was the
The turbidities of the supernatants are measured and plotted as
method. One outcome of his studies with the jar-test
a function of coagulant dosage. After plotting settled-water
apparatus was the technology of paddle-wheel floccula-
turbidity versus coagulant dose, another test is done with
tion, first implemented at the Sacramento WTP. Further,
smaller dosage increments, for example, to confirm an
by means of the apparatus, he was able to study not only
‘‘optimum’’ dose, and to obtain increased resolution. In add-
the effect of rotational velocity mixing on coagulation
ition, alkalinity may be added at an optimum coagulant dos-
and flocculation, but the effects of pH, dosage, alkalin-
age, then various polymers, etc. In one case (known to the
ity, etc. on the character of the floc produced. Black and
author), some 5000 jar tests were done by the consulting
Harris (1969, p. 49) confirmed the foregoing account,
engineer retained for plant modification (Box 9.3).
stating that the test was first used by W. F. Langelier and
Charles Gilman Hyde in 1918 to determine the proper
treatment for the water of Sacramento. The laboratory
BOX 9.3 ORIGIN OF THE JAR TEST
studies, they stated, led to the first large-scale applica-
A jar-test apparatus was developed and used by Profes- tion of ‘‘mechanical agitation,’’ that is, paddle-wheel
sor Wilfred Langelier (Langelier, 1921) who stated: flocculation (Chapter 11), which was at the Sacramento
WTP in 1921, as recommended by Langelier.
The apparatus used to determine the ideal conditions of In a later study, A. P. Black with two colleagues,
agitation consists of a wooded frame on which are Owen Rice and Edward Bartow (Black and Rice, 1933),
mounted four rotating paddles or propellers, each geared used six ‘‘battery jars’’ of 3.5 L capacity for their
to a horizontal shaft driven by a small electric motor. The coagulation experiments, with 2.0 L water used for
water is held in cylindrical glass jars 5 in. in diameter and each jar. The standard jar-test apparatus as an ‘‘off-
10 in. deep. A later modification of the device is
the-shelf’’ item was developed and marketed by Phipps
designed with friction disks which permit the operation
& Bird Inc. and was available by 1935.
of each paddle at a speed independent of the others. The
same effect of variable agitation can be produced in the
original apparatus by using paddles of different sizes.
These simple devices have proved extremely useful in 9.8.2 BENCH SCALE FILTERS
studying coagulation phenomena. Similar apparatus
could with profit be included in the laboratory equip- The standard jar test does not work well for some water, for
ment of all water purification plants using coagulation. example, low turbidity (e.g., <0.5 NTU), low alkalinity (e.g.,
<50 mg=L as CaCO 3 ) snowmelt waters found in the Rocky
The context was his work with Professor Charles Mountains, the Sierra Nevada Mountains, the Cascades, and
Gilman Hyde in the design of a water treatment plant other places. While microflocs form, as evidenced by the
at Sacramento, California (Langelier, 1982). The issue increase in turbidity after coagulation with alum in a jar test,
that Langelier was addressing was coagulation. At that a settleable floc does not develop as in higher turbidity waters.
For such situations, ‘‘in-line’’ filtration is appropriate