In weeks two and three of this course, we looked in some detail at the processes and mechanisms involved in different HWTS technologies. Each technology has its own strengths and challenges. And when HWTS technologies are being considered or selected, those should somehow feed into a decision making process. Selection of HWTS options can happen at different levels. At the household level an individual can decide to use or not to use a given technology. But frequently decisions on technologies are taken at a program level or even by central governments. And how can these decisions then be made in a rational way at these different levels? Well, the first thing to do is to be able to say what is wanted. What are the characteristics of an HWTS option that are important for selection? For application of HWTS in a particular context. And that involves clearly defining, and establishing the context both in terms of the water to be treated but also all those other factors, which are outside the control of an intervention like the policy environment, existence of supply chains, etcetera. Once the context is clear, a selection among potential HWTS options can be made by defining objectives or criteria. And then there's usual more than one criteria. So combining those in some way and assessing different options against those different criteria, and taking a decision. This can be done in a formal way through a multiple criteria decision analysis approach or informally and even subconsciously. Selection of an HWTS option is highly context specific. Of course, the context includes the characteristics of the water to be treated. An important factor here is whether or not the water is likely to be turbid. But also the level of contamination is important. Some processes can achieve higher LRV's than others. Apart from the water quality, the way that the water is assessed and the quantities available and needed can also influence selection. Then there are other contextual factors not related to the water, such as those described in the IBM WASH model, like the wealth and education of the target population, the access to markets, supply chains and information, the policy environment. And, to these, I would add nature of the supporting organization. Because a social business and a profit oriented business might choose very different options based on their differing objectives. Here are some examples of sets of criteria that have been put forward by various groups working on HWTS. There are many such examples and I just selected five here, that were fairly well documented from these organizations or research groups. Let's take a brief look at each of them. You've seen this WHO training manual by now, a few times hopefully. It really contains a lot of useful material. And in the section on selecting an HWTS options, the authors identify five sets of criteria shown here. First, effectiveness. How effective are the different HWTS options for the water to be treated in the given context? Next, appropriateness. These are also contextual factors, related to the conditions and desires of the target populations. Things like local availability and supply chains, the amount of time required for treatment, requirements for operation and maintenance. And the lifetime of the option. Appropriateness is closely related to acceptability. Which the authors takes to include aesthetic aspects of the treated water, it's taste, it's smell, it's color. But also perceptions of convenience, health gains, social status, time and money savings, etcetera. Of course, the cost is important, there are different types of costs. Upfront and recurring costs to the household. But also costs to the supporting agency, which may include subsidies, and an appropriate business plan. Finally, implementation is a bit of a catch-all class of criteria. It includes things like how the technology is manufactured and distributed to the households, how fast the technology can be implemented. What kind of monitoring is required for the technology. How well it can be integrated with current government programs, etcetera. I encourage you to go to the manual itself for a fuller discussion of these criteria. Here's another set of criteria. This one from Professor Mark Sobsey at the University of North Carolina, Chapel Hill, and colleagues, from a few years back. They came up with a set of five, sets of criteria that should be satisfied for HWTS to be sustainable. I let you just read them on the screen and refer to the paper for a more detailed discussion. The Netherlands Water Partnership has a nice brochure on HWTS. We can download it here. And, they propose a set of 13 indicators grouped into the categories of performance, people, and planets. Now, this is a bit different from the first two in that this category of planet considers brings in environmental considerations, as well. At EAWAG, I was part of a large project on mitigation of arsenic and fluoride, called Water Resource Quality. Some of you have asked for information about HWTS for improving chemical water quality. And it's a really interesting topic. But, unfortunately, beyond the scope of this course. You can get some good information from the Water Resource Quality website here. And one of the journal publications that came out of this project, that you can see here, describes a process for selecting a water treatment option. Even though the technology in question was for fluoride removal, the process could still be useful for selecting HWTS options. The author has identified 12 criteria. Grouped into three broad classes of Affordability, Acceptability and Reliability. One last example now. This is from the US NGO PATH, which implemented the safe water project, that did a lot of great work about commercialization, of HWTS in developing countries. Spend some time at their website, there's a wealth of information. In particular, they've established generic design guidelines for HWTS. If you ever find yourself building or designing an HWTS system, this is a really useful set of information. Some of the, highlights of this projects were summarized in a special report of PATH's Perspective magazine. You can download that here from their website, definitely worth the read. As part of the Safe Water Project, one of the PATH staff, Stephen Himley, completed his Master's degree in Public Health in the University of Washington. And his topic was a very detailed and rigorous examination of how to select an HWTS option for a given context. He described three main criteria, acceptability, availability, and affordability. And then he broke down acceptability into six sub-components, so that there are a total of eight criteria. Okay, so there are lots and lots of lists of criteria out there. So what? Again the important thing isn't to decide which set of criteria is the best or the most efficient, but to identify for a particular application of HWTS. Which criteria are the most important for selecting an HWTS option, and it might not be the same set everywhere. I hope that these examples can help you develop a set of criteria if you ever find yourselves in that position. And if you do, and you come up with a set of criteria, well, then what do you do with them? Well, there are different ways to assess each possible HWTS option against the selected criteria. This can be done in a fairly easy, non-quantitative way with landscaping, or through a more complex and formal optimization process. A slightly different approach is to use a decision tree. Posing a series of sequential questions that allow you to narrow down the list of potential technologies. Now, here's an example of landscaping. Again, from the PATH safe water project. It's their Technology Assessment and Ranking Tool or TART. Now please, don't try and read the text of it. It's just an example of landscaping. But it's a table which contains 18 criteria, or dimensions plotted against 15 different HWTS systems. And each of those systems is rated qualitatively, good, neutral or bad, for each of the dimensions. And then they are colored green, yellow, or red like the traffic light system. And in the landscape the different criteria are not combined or weighted into an overall score, but everything is displayed on this one dashboard. So you can see that a technology like this one, which is mostly green, or, maybe, this one with just one red is probably a better option. Than this one here, where you see lots of red or, or maybe this one. Now, landscaping doesn't select a particular optimal choice. And it could be that some of the criteria up here are more important than other ones. And you would need to consider that when interpreting a landscape matrix. One way to overcome that problem is to make more formal optimization with quantitative scoring, and weighting of different criteria. There are different frameworks for this with different names. This one here is Multi-Criteria Decision Analysis. But, the basic idea is usually something like this one. We've talked already about establishing a decision context. And of course you have to come up with a list of potential options for consideration. And then we've also talked about objectives or criteria. Once you have criteria, you can score each option each technology for each of the criteria using some objective system. And then weight the criteria so that the ones that are more important count more. Then by summing up the weighted stores, you arrive at a total score and then look at the results. It's also a really good idea to do a sensitivity analysis. To see if one particular criterion is dominating the selection process. Here is an example of scoring, again, from, Steven Himley's MPH thesis. And again, don't try to read the text, it's just an example that for each of these eight criteria here There are rules to assign a score. For example the top row here is the time to treat and if the technology takes more than 20 minutes to treat water you get a poor score. Whereas if it takes less than five minutes it gets an excellent mark. From that same thesis, here is an example of weighting. The three main criteria were assigned different weights, always summing up to 100%. For different contexts, urban and rural lower and low income settings. So, you can see that affordability is given the highest most important weight in a low income, rural context. Whereas acceptability is given the highest weight in a low income urban context. And this acceptability criteria, had six sub criteria, which were also given different weights for each of those four contexts. These weights then could then be combined with the scores and presented as a total score. Here you see total scores for two scenarios. On the right you have turbid water, and on the left, non-turbid water. You could have a different optimal choice, depending on the turbidity. So it looks like in this case chlorine is a liquid or tablet ends up being the best option in all of the scenarios. Now that might sound funny because chlorine has problems with turbid water. And what probably happened is that technical effectiveness here is only one of the criteria. And apparently didn't have a high enough weight to overcome the greater affordability in particular of chlorine compared to other options. If you can see, if you look just at the acceptability scores. That some of the non-chlorine options do a little better than chlorine in the turbid scenarios. But this example does highlight that a formal optimization process is tricky. And can sometimes come up with recommendations that might not seem intuitive or correct. A simpler approach to selection is the familiar decision tree. By asking a series of questions, you can narrow down the group of potential options appropriate for a particular context. Here's a decision tree adapted from that IFRC report on HWTS in emergencies, which you saw in module 4.3. Here, if the water is muddy, the tree says that straining, the 3-pot method, coagulation, and sedimentation or filtering with frequent cleaning should be promoted. And these technologies are all appropriate for turbid waters. If the water is not turbid the next question is, whether or not HWTS products are available from outside the community. If they're not, then you get directed towards boiling, if fuel is available. Or three pot methods, sedimentation, SODIS, and filtration, if fuel isn't. If HWTS products are available. Then the tree directs you towards chlorine. Either with a normal, or a double dose, depending on the turbidity. In all cases, safe water storage and handling is to be promoted. So that's one example of a decision tree for a particular context, that of HWTS in emergency response. It shouldn't be applied universally but it could be a good model for developing a context specific decision tree. That could be used by decision makers perhaps more easily than a full quantitative optimization process. So in conclusion, it's not so easy to say which HWTS option is best and there is certainly no silver bullet, no answer to apply in every situation. The the best solution will depend on the local context but also on the criteria that are considered to be the most important. And these criteria might be given different weightings, particularly if a formal optimization approach is applied. But there are also less formal approaches like landscaping and decision trees, that can be used to narrow down the list of options. In many cases, the goal should not be to arrive at one single option. But to narrow down a long list to a short list. So that people still have some freedom to select the option that works best for them in their particular household situation.
