More troubleshooting and some thoughts on how to solve problems

In between our participation in two seminars/tours and waiting on container repairs, we have continued to operate our system and attempt to get data. Also, we had a final report due to our funding agency (USAID ) on August 15.

We are continuing to have problems related to the particulates from the Los Almendros well. Particles, like sand and other things, are common with wells. The Los Almendros well is able to produce much more flow than what our pilot needs, so we are cycling the well multiple times per day to fill a tank which feeds our system.Wells aren't normally cycled like this, or, if they are, a sand strainer is installed prior to a membrane system. We didn't know about the particles prior to our arrival, so we don't have this level of pre-treatment. The particles seem to be causing problems in our the RO/NF and EDM systems (here's a link describing our process if this is the first post you have read). We will need to change out our cartridge filters much more often than what is normally done with a brackish well water pilot. This is a fair amount of work and Clara and I, try as we might, just can't get the housings loose all the time.

Cartridge filters and evidence of particles from the well. From top left, going clockwise:
1-Clara and I trying to get the cartridge filter loose (failed). 2-Bottom cartridge filter vs new filter. 3. Three filters showing 1 week's worth of particle filtration, 2 day's worth of particle filtration, and a brand new filter for comparison. 4. sample of the water in the cartridge filter housing with particles rejected by the filter. 5. Scale and particles in the EDM feed manifold.

The high level of alkalinity (Wikipedia describes it pretty well) present in the Los Almendros well has caused precipitation to occur without acid pre-treatment or a lot of water for dilutions. In previous pilots with high alkalinity, acid pre-treatment or substantial dilutions to the concentrate stream were implemented to prevent scale. We have attempted to operate without both of these because increased dilutions reduce the overall recovery and the Desal Prize and piloting objectives prioritized less chemicals. However, sulfuric acid is a common chemical and is relatively easy to procure and use safely. The minimum recovery with dilutions alone will be determined in the coming weeks. Acid pre-treatment will also be attempted to show the maximum recovery possible.

Water samples from pilot. Having solids precipitate is uncommon from the UTEP team's experience (especially precipitation within 4-6 hours!)

The high level of silica and higher than expected rejection by the RO/NF is causing scale in the RO/NF. It is possible that the particles (possibly iron, or sand with iron) from the well are enhancing this problem. We will test operations with a silica bleed (send a partial stream of RO/NF reject to drain) to see if the problem is mitigated. We cleaned the RO/NF and installed a bleed on the RO/NF to test whether it will help with reducing scale buildup. We will attempt to identify the minimum waste flow to protect the RO/NF and estimate the additional evaporation capacity necessary to maintain the zero liquid discharge aspect of the system.

Our container will be repaired over the next few days, so we will have our system off. We expect to be up and running for the next several weeks and hope to have good news to report based off our findings thus far.

Fixing Pumps and Making a Video

We have kept our system off because we thought our container was going to be repaired. Since they will need to use grinders, which will make sparks, we want to keep our batteries from charging (and releasing hydrogen). The EDM also produces hydrogen when we are desalinating, so everything has to be off for the repairs. We have had miscommunication about timing, but hopefully we can operate for the rest of the week to obtain data for our final report to USAID.

I meant to include a picture earlier of Mr. Rusbel. Occasionally, Clara and I find something that we can't do without a little help. As I mentioned in my troubleshooting post, we needed to install cartridge filters with smaller pores to protect our membranes from particles from the well. Clara and I were able to get one replaced, but it was leaking. But, we couldn't get the other one replaced because we couldn't get the housing loose. Mr. Rusbel, with his characteristic smile, jumped right in and (with Clara's assistance) tightened the loose one and replaced the other. He's a great guy.

Mr. Rusbel & the cartridge filters (the bottom one is the one we couldn't remove without his help)
(Photo Credit: Clara Borrego)

So, what did we do today? Well, first, we performed surgery. Some of the previous problems with previous precipitation in the EDM feed ended up causing our pump to fail last night. The stack is operating fine, but the pump wouldn't pump. The scale (precipitated salts) were causing friction on the pump. These are really cool pumps, by the way. They have a magnetic drive, so there are very few moving parts. Another cool thing about them is that they use DC energy, so we can use power directly from the sun - no need to convert to AC power! This and other modifications reduced our electrical load by at least 33%. We took apart the pump and scrubbed off the scale. It's working like a champ now!

Before (left) and After (right) pictures of EDM pump

While waiting for the repair guys to arrive today, Jose Eveline (UPi student) came by to get some more footage for a video he is preparing for our UTEP-UPi project. He already has some footage of the system operating and today was slated for interviews. It was a lot of fun. Jose is also the one that took us to get coffee on our first weekend here in Tegucigalpa. 

How to get a shot with a short interviewee (haha!)
(Photo credit: Clara Borrego)

Clara, the sound engineer 

Clara, being filmed