water mixture containing 6 to 8% oil by volume through the elements. 
The optimum water flow rates were 0.8 gpm through Element No. 1 
and 1.5 gpm through Element No. 2 respectively. The mixture jet 
reattachment distances from the corners were measured for both 
elements. They were found to be 7 inches for Element No. 1 and 
7.5 inches for No. 2. These values are very close to the theoretically 
computed values listed in Table 1. The test results indicated 
that a portion of the oil in the mixture jet did accumulate 
in the separation bubble zone of the flow. A photograph of 
the flow pattern (Figure 5) through Element No. 2 at its optimum 
flow rate clearly shows the accumulation of oil in the separation 
bubble. This oil when extracted contained about 50% water. 
Consequently, improved designs for collecting oil transferred 
into the separation bubble were sought. One such design is 
that of providing a chamber at the top of the separation bubble. 
This chamber is connected to the separation bubble by means 
of holes in the top cover plate of the element, see Figure 
6. Under optimum conditions, the oil captured by the separation 
bubble flows into the collecting chamber through the connecting 
holes. Next, the oil collected in the chamber is transferred 
by siphoning to an oil storage tank. Two different designs 
of the collecting chambers were tested. These are shown in 
Figures 7 and 8. Due to its shape and its greater depth, the 
collecting chamber design shown in Figure 8 is more efficient 
in collecting the oil. Tests conducted on the elements with 
modified design show that about 50% of the oil in the input 
flow can be extracted in this manner, whereas the remaining 
oil flows out with the attached water jet. Furthermore, the 
oil being extracted contained about 5Z water. Thus, a separating 
device based upon this concept appears to be capable of gross 
separation only. However, tests on elements with modified designs 
must be conducted before deriving final conclusions about the 
degree of separation obtainable. A photograph of the flow pattern 
through the Element No. 1 taken at its optimum flow rate is 
included as Figure 9. The accumulation of oil in the collecting 
chamber is clearly visible in this record. 
Tests On A Multi-Stage Element 
It was realized during the feasibility tests on the single- 
stage elements that to make the Coanda-effect separator suitable 
for practical applications, staging is necessary. The number 
of stages for the separator, however, depends upon the type 
of oily wastes being handled together with the quality of effluent 
desired. 
During the course of this study a three-stage test element 
was designed and built to evaluate the effect of staging. Figure 
10 shows the sketch of the element’s middle plate with flow 
passages cut in it. Each stage of the element has a 1/4 inch 
