Overlapping vocations
sorption science and agroecology
I just returned from a ten-day stint with friends and colleagues at Caminos de Agua in Guanajuato state, central Mexico.
I’ve been working with Caminos since 2016. Pre-covid I would spend a week or two with them, three or four times a year, to help out with their excellent research team. Since covid my visit frequency has dropped to once a year, but hopefully will pick back up in years to come.
During my previous visit in summer 2021, Caminos inaugurated a pilot groundwater treatment systems (GTS) serving a community on the outskirts of San Miguel de Allende. The pilot GTS included a bone char sorption contactor for removal of excess fluoride. This treatment step was the main focus of my recent visit.
Ribbon cutting ceremony, pilot groundwater treatment system, summer 2021.
For the past few years we have been working on a research project to (1) identify a minimalist approach to rapid, lab-bench sorption data collection, and (2) build a practitioner-oriented work flow for using sorption data in conjunction with modeling to support treatment system design and value engineering for using bone char to remove excessive fluoride.
Here’s the idea: imagine you’re a WASH practitioner sent on a mission to, for example, Tanzania. (Groundwater from the Rift Valley of east Africa is known to be high in fluoride.)
Elevated fluoride in groundwater.
The locals bring you a jerry can of groundwater from their well, and a sample of charcoal made from cow bones.
Time, money, and resources are limited. Can you perform one rapid lab-bench test and learn what you need to know to design household and community water treatment systems using bine char to remove fluoride to safe levels for drinking?
After my most recent visit to work with Caminos’ tech team, we can answer this question in the affirmative.
Using data collected over the past 2-4 years, we have developed a “standard” rapid small-scale column test (RSSCT) for evaluating bone-char/groundwater combinations of interest.
The test takes 1-2 days to complete, and provides quantitative information that can be used to:
Compare different bone char sorbents and quantitatively rank them by effectiveness for treating a particular groundwater
Generate simulations of full-sized systems that can treat water over periods of weeks to months, using data collected from experiments that last only a few to several hours
Explore operational conditions - contactor size and contact time, different levels of fluoride in influent water - to optimize contactor design, sorbent replacement frequency, and other parameters related to treatment system value engineering.
How did we do it?
For one, we took advantage of the fact that, compared with other water treatment circumstances such as adsorption of organic chemical contaminants in the presence of dissolved organic matter, the bone char-fluoride system is significantly simpler.
Typically, the predominant water quality factor influencing fluoride uptake by bone char is pH. As pH increases, fluoride uptake by bone char decreases. For the bone char-groundwater system, interference from competitive sorbets tends not to play a significant role compared with the effect of pH. (I write tends because there are always exceptions, and “your mileage may vary…” so use appropriate caution.)
It’s often the case (though not always of course) that groundwater throughout a region will be similar in pH and background water chemistry. This is the case for most of the community wells in the watershed served by Caminos de Agua. pH in this region is high - typically 8.0-8.5 - and fluoride levels mostly range from 1.5-6.0 mg/L (though a small percentage of wells contain fluoride in excess of 10 mg/L or even 20 mg/L).
Caminos de Agua’s groundwater quality monitoring map.
If pH and other background water quality parameters can be effectively regarded as constant across a region, then the two main factors affecting bone char treatment system performance are (1) influent fluoride level, and (2) contactor operational parameters, namely, contact time.
It would be costly, time consuming, and onerous to perform a major campaign of experiments to test every possible combination of fluoride concentration and (empty bed) contact time (EBCT) in the process of designing and value engineering treatment units.
This is why we have sought to develop a mass transfer modeling approach - one that can provide accurate simulations of fluoride sorption after training to a bare-minimum set of empirical data.
Computer models are great! You put numbers in, run the model computations, and voila! you get a simulation output - every time!
Now, whether than simulation output means anything, whether it has any relationship to reality, is another question altogether.
What I and the Caminos tech team staff (mostly the tech team staff, tbh) have been diligently working on the past couple of years is a research project to validate, if possible, a mass transfer modeling approach capable of producing simulations that match actual real-world empirical data.
Long story short, we did it!
Specifically, what did we do?
We determined equilibrium and kinetic parameters for fluoride sorption from Caminos’ groundwater using Caminos’ bone char. (I.e., very representative base-case conditions for using bone char to remove excess fluoride from groundwater.)
We used a common set of fitting parameters as model inputs, and fit the homogeneous surface diffusion model (HSDM; a common sorption mass transfer model) to RSSCT data collected using (1) different fluoride concentrations (2-10 mg/L), and (2) different EBCTs (10-90 minutes).
We obtained good fits of the HSDM to fluoride breakthrough data collected under this range of fluoride influent concentrations and EBCTs, which confirmed the validity of our modeling approach. The implication of this work is that
Accurate model simulations can be obtained using one set of experimental (RSSCT) data to train the HSDM to a bone-char/groundwater combination of interest.
This is a really cool result!
It means that, if you find yourself on that mission in Tanzania, you can run one RSSCT (which will only take 1-2 days) to determine key model inputs. Then you can use the cleverly designed freeware program FAST to run HSDM simulations and build scenarios for treating waters containing different concentrations of fluoride using contactors of various sizes (i.e., EBCTs).
This work will be hugely beneficial to Caminos as well as to WASH practitioners around the world. Accordingly, we’re working on peer-reviewed paper describing the results and outlining a workflow for field practitioners to use the method we developed. When that paper comes out I will share it here.
In addition to the brilliant and hard-working Caminos tech team, this project was made possible through building on the excellent work of my former MS student, Maggie Thompson. Maggie experimentally measured equilibrium and kinetic parameters for the bone char/fluoride system, and developed an RSSCT method for accurately simulating full-size contactor performance. You can read about her work in our previous publication here:
Thompson MP, Kearns JP. 2021. Modeling and experimental approaches for determining fluoride diffusion kinetics in bone char sorbent and prediction of packed-bed groundwater defluoridator performance. Water Research X, Vol 12., August 2021.
What’s this about “overlapping vocations”?
During my recent visit with Caminos I also got to tour a nearby agroecology project: the Via Organica ranch.
It is a stunningly impressive project. I took a few short video clips to give a sense of the place.
Among their initiatives is a project to plant 1 billion agaves all over Mexico.
The gardens are immaculate.
Friendly burros!
Innovative rabbity system.
In addition to low-tech water treatment, I have a major passion for agroecology and regenerative agriculture. It was great to be able to fuel both of these passions on my recent trip to Central Mexico!
I’ve been writing another Substack newsletter about the agroecology/regen ag projects that my wife and I are working on at our place and with friends in our bioregion of Southern Appalachia.
This newsletter is called Magpie Hollow Farm News, and it is 100% free to sign up! Please check it out!
Alabama (rear) and her boys born January 2022 - Gulab (right) and Jamun (center).
One of the features of the newsletter is a photo-essay series entitled “Then & Now” which shows through photos and a little text description the changes and developments taking place on our farm since we moved here a little over two years ago.
We hope you enjoy it!