Experimental Method for Retrofit Designs

One of the contributing factors to air entrainment at the beginning of the water treatment process in AguaClara plants is the waterfall effect inside the Linear Flow Orifice Meter (LFOM). This effect is composed of two basic entrainment mechanisms: 1) the surface of the jet of water becomes sufficiently turbulent to entrain air before it reaches the surface water; and 2) the penetrating jet creates a void from which small air bubbles pinch off. To minimize the effect of these entrainment mechanisms the following retrofit design options were calculated and tested in lab:

1) Pipe Inside LFOM: A pipe of a small diameter would be placed within LFOM with about a few inches of separation between the outer surface of the pipe and the inner surface of the LFOM.
2) Vertical/Inclined Plane: A vertical or inclined plane would be placed within the LFOM that would prevent the water jets from hitting the surface of the water directly.
3) Bucket with Holes: A bucket would be placed inside the LFOM that would catch the water jets and slow down the inflow of water by using an orifice.
4) Bucket without Holes: Rather than have an orifice slow down the flow of water in the LFOM, the bucket would be over flooded so that the water dribbles down the side of the bucket.Once our team figured out the cause of the foam forming in AguaClara plants in operation in Honduras (put link citing causes), we set about designing retrofits for these plants. Our team came up with four different designs to test:
-the "pipe within the LFOM"
-the vertical/inclined plane
-the "teacup" theory, using an overflowing bucket to catch the falling water
-a variation on the teacup theory, using a bucket with an orifice in the bottom

See the retrofit designs page for visuals of each theoryoption.Our first goal was to narrow down from 4 different designs to our top choice to facilitate and expedite lab testing.

In addition to these designs, we also tested for the three limiting parameters of foam formation from water jets. You can find the summary of this experiment at the Experimental Method for Limiting Parameters of Foam Formation from Water Jets page

Calculations

As a team, we worked in MathCAD to calculate the distance that the jets of water coming into the LFOM would travel inside the LFOM. If we discover had discovered that a bucket inserted into the LFOM below the orifices can could catch the jets before they hit the water, the bucket theory will would be a viable option. We determined that all but the top three jets of water would in fact hit the far wall of the LFOM before reaching any size bucket that we could place in the LFOM. This Also, we used MathCAD to determine the size of the orifice needed in the bottom of the bucket to maintain the plant flow rate. We determined that the orifice would have to have a 4.5cm radius, which was far too large for our buckets. The MathCAD calculations eliminated both the teacup bucket theory and the variation on that bucket with a hole theory as possible solutions. The MathCAD file is files are attached. (what was the purpose of doing this?)
(Please document equations used and summarize results obtained on the wiki as well)
Jet Distance Calcs 7-13-09.xmcd
CDC_Retrofit_Designs.xmcd

Testing

After eliminating both bucket theories, we decided that testing the last two designs in a lab would not be difficult and would give more realistic, tangible results than our calculations.

...

We used the pilot plant to test our retrofit design in an environment that replicated the actual plants as closely as possible. We set the flow rate of the pilot plant at about 150L/min, which, coupled with the 7.5cm-diameter LFOM, would be roughly comparable to the 400L/min, 15cm diameter LFOM in the plant at Cuatro Comunidades.

The first design we tested was the vertical/inclined plane. We used a 5cm-wide, 60cm-long steel plate to test the effectiveness of this plan to reduce water aeration. We were looking at the amount of bubbles formed under the LFOM to determine results, and documented them photographically.

Next, we used various-sized pipes inserted within the LFOM to test the "pipe within the LFOM" theory. First, we tested a 3cm-diameter pipe by inserting it vertically into the 7.5cm-inner diameter LFOM. Second, we tested a 5cm pipe, and noticed that it produced a reduction of bubbles. After leaving the 5cm pipe in the LFOM for 5 minutes to ensure no overflow, we moved to test the 6cm pipe. This pipe also showed signs of reducing the amount of bubbles, so we left it in for 5 minutes as well to determine if it would overflow, which it eventually did.

Results and Discussion

Vertical/Incline Planes

Unfortunately, we determined visually that there was very little, if any, change in the amount of bubbles formed before and after inserting the steel plate into the LFOM. Even at various angles of inclination, the amount of bubbles formed was the same. We believe it was ineffective for two reasons. First, the spread of orifices wraps around the pipe far enough that the jets of water at the edge do not touch the steel plate. Second, the water jets are flowing fast enough that the water simply ricochets off the steel plate, instead of running down the plate as intended. See the gallery of photos below for photos illustrating our results.

"Pipe within the LFOM"

With our initial test of the 3cm pipe, there was no noticeable change in the amount of bubbles formed under the LFOM. For the 5cm-diameter pipe we found a drastic reduction in the amount of bubbles formed under the LFOM. Keeping the pipe in for 5 minutes showed no signs of overflowing the LFOM or even raising the water level, so we can assume that 5cm is a possible solution. Finally, we tested a 6cm pipe for 5 minutes in the LFOM. We noticed that the water level outside the LFOM rose slowly throughout the 5 minutes until the water flowed over the top of the LFOM. This clearly meant that the 6cm pipe was too big, and was causing a backup of water within the LFOM. (attach mathcad sheet here)

Vertical Plane and "Pipe within the LFOM" tests and results can be found at the vertical plane page and the "pipe within a pipe" page.

Results and Discussion

We We were able to document results for each variation of the designs visually with photographs and short movies on cameras, see the gallery below to view the photos.

Conclusions

At the end of all of our testing, we determined found that inserting a vertical plane into the LFOM would not effectively reduce the bubble formation enough to solve the problem. However, we did determine that inserting a 5cm-diameter pipe would effectively reduce the amount of bubbles produced by the LFOM, while not constricting the overall plant flow rate. This would translate to about a 10cm-diameter pipe in the AguaClara plant at Cuatro Comunidades. Hopefully this design change can be easily executed in the AguaClara plants in Honduras to fix the foam problem.