Discussion by Daniel A. Okun
Professor of Environmental Engineering
University of North Carolina at Chapel Hill
Having initiated laboratory studies utilizing pure oxygen for wastewater treatment twenty-five
years ago (Okun, 1948), I am naturally pleased with the renewed interest in it. The original
experimental investigation with oxygen evolved in the so-called "bioprecipitation process" (Okun,
1949). This process represents a substantial departure from the conventional activated sludge process.
The pure oxygen is dissolved in the influent flow to an upflow reactor unit. The biological floe
(activated sludge) in the reactor draws oxygen from the influent flow. In operation, the reactor
effluent contains residual dissolved oxygen, signifying that dissolved oxygen is present throughout the
sludge blanket in the reactor.
The study demonstrated that:
1. The efficiency of the process is a function of the weight of volatile solids (biological floe or
activated sludge) in the reactor, and the process of preoxygenation of the influent permits much
higher concentrations of volatile solids in the reactor than was then possible in a conventional aeration
unit.
2. Because of the higher solids concentration that can be maintained in the reactor as compared
with the aeration tanks of conventional activated sludge plants, and because the reactor combines the
functions of both aeration and final sedimentation tanks in conventional plants, this process was
estimated to permit a reduction in overall treatment system tank capacity to about 25% of that
necessary in conventional treatment, for the same overall removal.
3. The high dissolved oxygen concentrations maintained in this process prevented the excessive
growth of filamentous organisms that often occur in activated sludge treatment of municipal
wastewaters. When highly carbonaceous wastes were added to the wastewaters treated, filamentous
forms did develop, but because of the absence of vigorous agitation in the reactor, these presented no
problem to the treatment process and, in fact, resulted in a highly polished effluent.
The results were sufficiently attractive that a 1 0 - to 30-gpm pilot plant with a 4-ft diameter,
12-ft deep reactor was constructed at the Back River Sewage Treatment Plant in Baltimore in 1953.
The results for this pilot plant were, in turn, sufficiently promising that a larger plant, 12-ft diameter
and 22.5 ft deep (Fig. 1) designed for flows of 50 to 75 gpm, was built in Stamford, Connecticut in
1955 (Budd and Lambeth, 1957). The only difference between this flow diagram and the original was
in the utilization of recirculated flov from above the sludge blanket to increase the weight of dissolved
oxygen that could be carried into the reactor. Influent to the plant could be introduced either before
or after oxygenation. Flow to this plant was varied according to the influent to the Stamford plant.
The clear liquor influent to the reactor was preoxygenated in a counter-current column, with oxygen
absorption efficiencies ranging from 30 to 80%.
The most significant conclusions from these studies were:
1. The operating power requirements for tonnage oxygen in this process and for air in conventional
activated sludge treatment were essentially the same for the same degree of treatment, and
2. Because the upflow unit serves as both aerator and final clarifier, the secondary treatment area
requirements can be reduced by 50% with the volume requirements for treatment being reduced by
about 30%.
Despite the promising results of these studies, interest in pure oxygen in municipal treatment
plants, and because demonstration grants, or "risk" funds were not available to induce a community
to invest in unorthodox treatment.
In 1968, the Union Carbide Corporation, partly because of its interest in selling oxygen and
oxygen-manufacturing plants through its Linde Division, introduced the Unox process. This process
uses pure oxygen in place of air in the conventional activated sludge process. In order to obtain
efficient utilization of the oxygen added, the aeration tanks need to be covered, and oxygenation is
performed in sequential chambers with the waste gases moving concurrent with the flow. With Federal
demonstration grant support Union Carbide Corp. operated a large scale demonstration plant at
Batavia, N.Y., which indicated savings on the order of 10 to 20% over activated sludge using air
(Albertsson et al, 1970).
Consulting engineer's evaluations of the Unox process as compared with conventional activated
sludge treatment generally conclude that Unox does not appear to offer significant economies for new
plants. However, the process does appear to be attractive where existing activated sludge plants need
to be increased in capacity and land area for plant expansion is limited.
The Open-Tank Pure Oxygen System
The work reported here by Cohen (1971) is a further development in the use of pure oxygen. If
the pilot plant results described are borne out in full-scale practice, the work represents a significant
401
Professor of Environmental Engineering
University of North Carolina at Chapel Hill
Having initiated laboratory studies utilizing pure oxygen for wastewater treatment twenty-five
years ago (Okun, 1948), I am naturally pleased with the renewed interest in it. The original
experimental investigation with oxygen evolved in the so-called "bioprecipitation process" (Okun,
1949). This process represents a substantial departure from the conventional activated sludge process.
The pure oxygen is dissolved in the influent flow to an upflow reactor unit. The biological floe
(activated sludge) in the reactor draws oxygen from the influent flow. In operation, the reactor
effluent contains residual dissolved oxygen, signifying that dissolved oxygen is present throughout the
sludge blanket in the reactor.
The study demonstrated that:
1. The efficiency of the process is a function of the weight of volatile solids (biological floe or
activated sludge) in the reactor, and the process of preoxygenation of the influent permits much
higher concentrations of volatile solids in the reactor than was then possible in a conventional aeration
unit.
2. Because of the higher solids concentration that can be maintained in the reactor as compared
with the aeration tanks of conventional activated sludge plants, and because the reactor combines the
functions of both aeration and final sedimentation tanks in conventional plants, this process was
estimated to permit a reduction in overall treatment system tank capacity to about 25% of that
necessary in conventional treatment, for the same overall removal.
3. The high dissolved oxygen concentrations maintained in this process prevented the excessive
growth of filamentous organisms that often occur in activated sludge treatment of municipal
wastewaters. When highly carbonaceous wastes were added to the wastewaters treated, filamentous
forms did develop, but because of the absence of vigorous agitation in the reactor, these presented no
problem to the treatment process and, in fact, resulted in a highly polished effluent.
The results were sufficiently attractive that a 1 0 - to 30-gpm pilot plant with a 4-ft diameter,
12-ft deep reactor was constructed at the Back River Sewage Treatment Plant in Baltimore in 1953.
The results for this pilot plant were, in turn, sufficiently promising that a larger plant, 12-ft diameter
and 22.5 ft deep (Fig. 1) designed for flows of 50 to 75 gpm, was built in Stamford, Connecticut in
1955 (Budd and Lambeth, 1957). The only difference between this flow diagram and the original was
in the utilization of recirculated flov from above the sludge blanket to increase the weight of dissolved
oxygen that could be carried into the reactor. Influent to the plant could be introduced either before
or after oxygenation. Flow to this plant was varied according to the influent to the Stamford plant.
The clear liquor influent to the reactor was preoxygenated in a counter-current column, with oxygen
absorption efficiencies ranging from 30 to 80%.
The most significant conclusions from these studies were:
1. The operating power requirements for tonnage oxygen in this process and for air in conventional
activated sludge treatment were essentially the same for the same degree of treatment, and
2. Because the upflow unit serves as both aerator and final clarifier, the secondary treatment area
requirements can be reduced by 50% with the volume requirements for treatment being reduced by
about 30%.
Despite the promising results of these studies, interest in pure oxygen in municipal treatment
plants, and because demonstration grants, or "risk" funds were not available to induce a community
to invest in unorthodox treatment.
In 1968, the Union Carbide Corporation, partly because of its interest in selling oxygen and
oxygen-manufacturing plants through its Linde Division, introduced the Unox process. This process
uses pure oxygen in place of air in the conventional activated sludge process. In order to obtain
efficient utilization of the oxygen added, the aeration tanks need to be covered, and oxygenation is
performed in sequential chambers with the waste gases moving concurrent with the flow. With Federal
demonstration grant support Union Carbide Corp. operated a large scale demonstration plant at
Batavia, N.Y., which indicated savings on the order of 10 to 20% over activated sludge using air
(Albertsson et al, 1970).
Consulting engineer's evaluations of the Unox process as compared with conventional activated
sludge treatment generally conclude that Unox does not appear to offer significant economies for new
plants. However, the process does appear to be attractive where existing activated sludge plants need
to be increased in capacity and land area for plant expansion is limited.
The Open-Tank Pure Oxygen System
The work reported here by Cohen (1971) is a further development in the use of pure oxygen. If
the pilot plant results described are borne out in full-scale practice, the work represents a significant
401