Economic Plating Waste Treatment by Reverse Osmosis
571
effect on a subsequent plating step, for example, chromium plating.
A preliminary cost estimate for reverse osmosis recovery of plating chemicals, based on
the laboratory results, indicated favourable economics. It was recognised, however, that
pilot plant evaluation is a necessary prelude to plant application, both to evaluate long
term operational effects and to obtain more accurate cost and design data.
FIELD TRIAL ON AN INDUSTRIAL NICKEL PLATING LINE
A field trial was commenced early July, 1971, on a small industrial automatic plating
line located in a factory in the Toronto area. The existing nickel plating operation utilizes
a semibright Watts nickel bath, maintained at 140°F, with evaporative losses of ca. 150
gal/day. Dragout rates vary according to the number and shape of the piece parts being
plated, but are generally of the order of ca. 15 gal/day. Three rinse tanks follow the
nickel bath.
Operated as a 3-stage countercurrent system and processing the rinse water from the
first rinse tank by RO it was anticipated that 99+ percent of the nickel dragout losses
could be recovered, and, depending on the amount of permeate that could be recycled,
deionized water usage might be greatly reduced.
It was recognised, however, that permanent installation of an RO recovery unit might
be uneconomical in relation to this particular plating operation. Prior to the RO field
trial, the three rinse tanks were arranged, as is typical of many small plating shop
operations, as follows :(1) A still rinse tank, filled initially with deionized water and rising from 0-3 oz/gal
before return to the bath as make-up.
(2) & (3) Two rinse tanks arranged in countercurrent, with ca. 1500 gal/day of deionized
water entering the final rinse tank, and flowing as effluent to the acid waste drain from
the second. Deionized water is favoured for quality rinsing, with minimum staining or
spotting of the plated articles.
Operating in this way, about 85% of the dragout nickel losses are recoverable.
The objective of the field trial was to obtain sufficient data over a test period of
several hundred hours to predict the feasibility of using RO to recover nickel values on a
larger scale, i.e. where nickel dragout losses are significantly high.
For several reasons, a commercial unit of the spiral wound configuration was selected
for the field trial. After preliminary studies in the laboratory, using a single spiral wound
module (50 ft
2 cellulose acetate surface area), the principal engineering parameters were
determined at 300-600 psi, and a compact, commercial mobile unit of this type was
procured, of capacity rated for the field trial. The unit has 100 ft
2 of cellulose acetate
membrane (2 X 50 ft
2 modules), with a nominal permeation rate of 1000 gal/day at 600
psi.
After design modifications, the RO unit performed satisfactorily during preliminary
trials in the laboratory, and was later installed on-site in the plating shop. An operating
pressure of 450 psi was selected as optimum.
A schematic flow diagram of the RO nickel recovery system is given in Fig. 2. The data
presented are typical of measurements made during intermittent operation over a
3-month period on a one-shift-per-day basis. Concentrations given are for nickel only. The
boric acid rejection was found to vary appreciably during the test for reasons yet to be
determined; however, an average rejection efficiency of ca. 50% for boric acid can be
taken as representative.
571
effect on a subsequent plating step, for example, chromium plating.
A preliminary cost estimate for reverse osmosis recovery of plating chemicals, based on
the laboratory results, indicated favourable economics. It was recognised, however, that
pilot plant evaluation is a necessary prelude to plant application, both to evaluate long
term operational effects and to obtain more accurate cost and design data.
FIELD TRIAL ON AN INDUSTRIAL NICKEL PLATING LINE
A field trial was commenced early July, 1971, on a small industrial automatic plating
line located in a factory in the Toronto area. The existing nickel plating operation utilizes
a semibright Watts nickel bath, maintained at 140°F, with evaporative losses of ca. 150
gal/day. Dragout rates vary according to the number and shape of the piece parts being
plated, but are generally of the order of ca. 15 gal/day. Three rinse tanks follow the
nickel bath.
Operated as a 3-stage countercurrent system and processing the rinse water from the
first rinse tank by RO it was anticipated that 99+ percent of the nickel dragout losses
could be recovered, and, depending on the amount of permeate that could be recycled,
deionized water usage might be greatly reduced.
It was recognised, however, that permanent installation of an RO recovery unit might
be uneconomical in relation to this particular plating operation. Prior to the RO field
trial, the three rinse tanks were arranged, as is typical of many small plating shop
operations, as follows :(1) A still rinse tank, filled initially with deionized water and rising from 0-3 oz/gal
before return to the bath as make-up.
(2) & (3) Two rinse tanks arranged in countercurrent, with ca. 1500 gal/day of deionized
water entering the final rinse tank, and flowing as effluent to the acid waste drain from
the second. Deionized water is favoured for quality rinsing, with minimum staining or
spotting of the plated articles.
Operating in this way, about 85% of the dragout nickel losses are recoverable.
The objective of the field trial was to obtain sufficient data over a test period of
several hundred hours to predict the feasibility of using RO to recover nickel values on a
larger scale, i.e. where nickel dragout losses are significantly high.
For several reasons, a commercial unit of the spiral wound configuration was selected
for the field trial. After preliminary studies in the laboratory, using a single spiral wound
module (50 ft
2 cellulose acetate surface area), the principal engineering parameters were
determined at 300-600 psi, and a compact, commercial mobile unit of this type was
procured, of capacity rated for the field trial. The unit has 100 ft
2 of cellulose acetate
membrane (2 X 50 ft
2 modules), with a nominal permeation rate of 1000 gal/day at 600
psi.
After design modifications, the RO unit performed satisfactorily during preliminary
trials in the laboratory, and was later installed on-site in the plating shop. An operating
pressure of 450 psi was selected as optimum.
A schematic flow diagram of the RO nickel recovery system is given in Fig. 2. The data
presented are typical of measurements made during intermittent operation over a
3-month period on a one-shift-per-day basis. Concentrations given are for nickel only. The
boric acid rejection was found to vary appreciably during the test for reasons yet to be
determined; however, an average rejection efficiency of ca. 50% for boric acid can be
taken as representative.