As discussed in our previous blog post, we planned to rerun the tests this week. However, we ran out of salt, so we had to place an order and are currently waiting on its arrival. In the meantime, we measured our O-rings to get them replaced, set up the feed tank with the new heater.

This week we also noticed a leak from the sediment filter, so we thought it would be best to replace the filter. We had a technician help us out so we wouldn’t ruin the piping, and it turned out that the sediment filter did not have a filter at all. We are unsure to what extent our fouling data will be affected, but we will rerun a “good dataset,” perhaps with the PTFE 0.45 micrometers membrane, and compare the new results for the one-hour and 4-hour tests with previous results. We noticed that it was still leaking after, so we sealed the inlets and outlets of the sediment filter’s casing with Teflon tape and pipe thread sealant because the water was coming out from that area. Afterwards, there was no leaking from that area.

An instructional document was made for the next cohort of students in the desalination team that might run into similar maintenance issues with the sediment filter. Linked here: https://docs.google.com/document/d/1DfIzhpXjXpY_DWhyBXxYKcoE5Elx8J7ssYWTODTqS3Q/edit?usp=sharing

For the rest of our time on Mountaintop, we finalized our methods for calculating head losses and head for our new future setup. Initially, our process flow diagram looked like below.

Figure 1 (original design proposal for heat exchanger)

We decided it would be better if we did calculations together instead of splitting up the work so that each of us calculated the pump’s required head consistently to ensure our calculations were consistent. So we calculated the pressure head of pumps 1 and 2 and considered any head losses by splitting up each cycle to determine the final pressures of each cycle and head losses, as shown in Figure 2. Note: Our flow calculations underwent multiple iterations as we tried thinking of a uniform way to calculate head, which is why our chart might look different compared to last week’s chart.

Figure 2

After meeting with Romero and showing him our updated process flow diagram and calculations, he suggested another heat exchanger to keep the permeate tank cool and to create a permeate cycle instead of a one-way flow to cool the new distilled water effectively. The new process flow diagram is shown in Figure 3. We intend to finish calculating heads for pumps 3 and 4 next week.

Figure 3

This week we finished running our last 7-hour test with the PVDF 0.45 micrometers membrane to complete our test matrix and meet with Dr. Romero to discuss the results. An issue arose because the heater was no longer maintaining the feed tank at 80 degrees Celsius around the last hour of the test. So we intend to run that test again.

The results from completing the test matrix earlier this week are shown below in Figure 1.

Figure 1

During our meeting, we made the Excel graph above and essentially decided to rerun some tests because the first-hour point for PTFE 0.22 micrometers membrane is below the 4-hour point, so we think there must have been an inconsistency while running the test. The other points we are rerunning are the 4-hour point for the PVDF 0.45 micrometers membrane and the 1-hour point for the PTFE 0.22 micrometers membrane. We are unsure why the points aren’t decreasing linearly for the PVDF 0.45 micrometers membrane and the PTFE 0.22 micrometers membrane, so we are rerunning these tests in particular. From there, we will choose the best membrane based on slopes, costs, and long-term effectiveness.

We also discussed with Dr.Romero how to make a flow diagram in Excel and have started making assumptions and calculations for the week. After we are done, Dr. Romero suggested running it through Bob Swan, our technician, so that he can make suggestions for valves and corrections. We continued to ask about the O-rings from the membrane’s casing because we wondered if it was normal for them to lift. He suggested we also have Bob Swan inspect them with a micrometer and replace them if needed. We also discussed potentially running the bubble test soon to check for leaks but were advised to wait until Monday because Dr.Romero would have guests on Friday, and he did not want a mess. We soon realized after the meeting that the direct contact membrane Desalination system’s sediment filter was leaking because there was a little puddle underneath it. We continued to work on our check-in presentations after following up with Bob.

After check-in presentations, we tried running our 1-hour test redo for the PTFE 0.22 micrometers membrane. However, we soon realized that the heater was not working. So we tried looking up a menu for the heat exchange and stumbled upon reviews mentioning that some customers have had issues with their heaters burning out and corroding. We tried unplugging the heater and the temperature controller, then plugging it back in, but that didn’t work. So we decided to inspect the feed tank and remove its insulation. We discovered the heater no longer worked because of corrosion, as shown in Figure 2 below.

Figure 2

The corrosion made it through the heater’s coils, which is why it no longer worked.

We then cleaned the feed tank by washing it with soap and water. We let it soak in water and residual dish soap over the weekend. We also ordered a new heat pump and should have it by Tuesday at the latest.

We continued to work on the process flow diagram and made some assumptions, like using a non-rigid clear PVC pipe for the heat exchanger cycle. Below are some screenshots of our work in Excel. Adam also created a Colebrook equation solver to determine the friction factor for our system( Figure 4).

Figure 3

Figure 4

We troubleshooted with the energy research center’s technician, Bob, this week. He helped us determine that the source of bubbles in the tubing was most likely originating from the membrane case. We were not sure if bubbles were supposed to form at the membrane. So we discussed it with Dr.Romero (when he returned from Mexico) later in the week if this is normal because one of the doctorate students is concerned about it ruining the pump. He essentially said that bubbles are normal, but to brush a dish detergent and water solution along the exterior of the system. If bubbles form, that means there is a leak. Otherwise, we might need to consult with Dr.Romero again and consider taking apart the feed tank.

During our discussion with Dr. Romero, we learned more about how to input information into a process flow diagram to make it more professional. We need to export our design to Excel and adjust our diagram from last week so that it is more applicable to real life. For instance, we cannot dispose of the system’s heavily concentrated salt water waste directly into the ocean (Figure 1) because it could hurt marine life, or pour it directly into the ground because it could hurt another ecosystem. So we would have to redirect it to a wastewater treatment system (Figure 2). Additionally, water has to cycle between the chiller and permeate tank to maintain the permeate water’s cool temperature. The permeate tank’s water cannot just go through the chiller and flow through the membrane channels, as shown in Figure 1. Below are pictures of the before and afters of the process flow diagram.

Figure 1 (below)

Before, our process flow diagram was very hard to read.

Figure 2 (below)

The after diagram is easier to read and follow along. We still need to adopt this into Excel and add all our known values (e.g., flow rates, temperature). For the time being, this is a much more accurate representation of a direct contact membrane desalination system.

We also troubleshooted this week on Wednesday when the power strip connected to the feed pump (Figure 3, above the purple insulated tank) and the temperature controller for the feed tank didn’t turn pn. We found this out after trying to heat up the feed tank early in the morning to prepare for a seven hour test for the PVDF 0.45 micrometers membrane. We had to contact the electricians because we weren’t sure why only that power strip wouldn’t turn on. It turned out that we were overloading that power strip in particular, so after they left we searched the energy research center for an extension cord to connect the pump to another outlet across the room, and connected our heater directly to an extension cord that was part of the power strip because the power strip did not work all together.

Figure 3 (below)

 Direct contact membrane deslanitation lab setup

The electricians inspected the rest of the system and pointed out that another power strip could be a one in a million fire a hazard because it is very outdated without a surge protector, it’s not meant for industrial purposes, and to check if we were overloading it as well and if possible replace it with a newer model. So we consulted with Romero and will be getting a new one next week.

Instead of running the seven hour test for the PVDF 0.45 micrometers membrane, we decided to just run the four hour test for the PTFE 0.22 micrometers membrane.

After our weekly check-in for the week, the Blockchain Systems and Applications complimented our presentation for being easy to follow. It was very meaningful to us because their expertise is distinct from ours, so we were glad they could understand what was happening.

Moving forward, we hope to finalize our process flow diagram before the end of next week so that we can start looking into purchasing equipment for our heat exchanger cycle and studying how hot tinaco water can get here compared to the Philippines. We also need to start building our network within the Philippines so that we can meet with them during fieldwork and perhaps even run a simple experiment. We would like to find out the maximum temperature that tinaco water can reach in the Philippines. Perhaps we could ask someone in advance to study the temperature of the water in their tinaco to validate our calculations and studies.

This week we developed a new test matrix in which we will be testing fouling. Fouling is the buildup on a membrane. Previously, we ran 

tests with the Direct Contact Membrane Desalination system to analyze how flux was affected by salinity level, the feed’s Reynold’s number, and feed tank temperature. Our new tests maximize flux based on the data we had collected in the spring, and the new variables are time and membrane material. So we will test the 3 membranes available to us, and run about 3 tests for each. The first test for each will run for an hour, the second for two hours, and the last for about 8 hours.

So far we have run 3 tests with the PTFE 0.45 micrometer membranes. It looks like the longer the tests are run, more loss is experienced. 

The above membrane is a result of the first test. Below is a picture of the membrane after the second test.

Above is a picture of a membrane after the third test which ran for 7 hours.

We also noticed that there are bubbles within the piping of our system and are getting a technician to come in, and figure out why. Bubbles can damage the pump and lead to inaccurate flowmeter readings.

While running tests, we are looking into developing another cycle with a heat exchanger and tinaco to warm up the feed tank with less energy requirements. As shown below.