Research findings from my thesis work

In my last post, where I shared my thesis, I promised you a short summary of my research findings. Here they are. (Well, some of them, anyway.)

First, a quick reminder of the research issue: In eastern India, ~85% of farmers irrigate their land using fossil fuel-powered pumps because they do not have access to electricity. Diesel and kerosene pumps have very expensive operation costs that farmers cannot afford, so they often choose not to irrigate outside monsoon season. With few other livelihood options, they migrate elsewhere for work. When they migrate seasonally, their families often cannot access social services such as health and education. I sought to address this challenge by developing an alternative off-grid pumping solution.

(Please do not confuse this with the opposite irrigation problem in western India, where free or near-free electricity for agriculture has led to unlimited pumping and severe over-exploitation of groundwater resources. The water table is not falling at a dangerous rate in eastern India. In fact some people argue that increasing groundwater extraction would reduce the incidence of flooding during monsoon season, because the soil would not be as saturated.)

For more about irrigation economics, see an earlier post here.

And now for my favorite research findings:

1. Rental costs are greater than fuel costs. This one actually surprised me. Everybody loves to talk about dirty fossil fuels and how the high cost of diesel (or in this case, kerosene) is what's keeping poor farmers from irrigating their fields. But I tested the actual performance of some pumps owned by farmers in Gumla and found that the flow rates were so bad that farmers needed to rent the pumps for an absurd number of hours to soak their fields--and hourly rental rates are fairly high. The implication? The focus of designing a new pump should be on increasing the flow rate more than fuel efficiency. The faster the flow rate, the less time it takes to irrigate a field--and the fewer hours required to irrigate, the less a farmer spends on renting a pump.

costs to irrigate one acre of land with 2" of water with a 12 year old Honda pump

2. Eliminating, or at least significantly reducing, suction head can reduce operation costs by up to 44%. I ran an experiment to test the hypothesis that eliminating suction head would increase efficiency and flow rates--and the experiment verified this hypothesis (see chapter 2 of my thesis). I then spent a great deal of time during my master's research trying to come up with affordable ways to run a submersible pump with a surface engine (see chapter 3 of my thesis), only to be outsmarted by the farmers (as usual): see #3.

3. Farmers already know #2. It turns out that farmers have already figured out that reducing suction head increases the pump's efficiency and flow rates (apparently, they experiment with lowering their pumps once the groundwater level falls beyond 7 m, the suction capacity of a pump, during dry season). Some lower the pump into the well with a rope, while others dig a trench next to the well:



4. Indian pumps are super leaky, but farmers know how to deal with it. I couldn't get an Indian pump to suck. Period. I tried a million ways to plug the leaks and just couldn't do it. That is, until I met a bunch of farmers who laughed at my teflon tape and showed me how it's done: slash an old bike tire and wrap those strips of rubber around all pump connections and hose fittings. Jugaad at its best! (But seriously, this is a problem the Indian pump industry desperately needs to address.)

5. Farmers claim Chinese pumps are more efficient than Indian pumps. Nobody knows why or if this is even true. Unfortunately I was unable to test a Chinese pump (I couldn't get them to work at MIT, and the villages I visited didn't use them--I would have had to go to West Bengal, and I did not have time). By taking them apart, I learned that they use a 3" discharge hose, rather than the 2" hose that Indian pumps use, and they have a smaller impeller-to-chamber volumetric fill ratio. The larger hose diameter makes sense to me: larger diameter = fewer pipe losses = higher efficiency. But I'm not sure how impeller-to-chamber volumetric fill ratio might impact efficiency, and I would be interested in exploring this issue further in the future (especially because I'm fairly certain Chinese pumps are going to take over the eastern Indian pump market).

6. Women don't touch engines. This one is not surprising. I heard from both men and women in the villages that "machines are for men" (though one women did pull me aside and asked me to show her how to start a pump engine when her husband wasn't looking). Women came up to me while I was testing the performance of their husband's pumps and asked, basically, "What about us? We're stuck with a bucket and rope. Our husbands may use a greater volume of water with their 'paani ki machine,' but we use water more frequently." To be honest, I had considered domestic water supply a completely separate issue from agricultural water supply, and I thought improving irrigation would increase the wealth of the entire family and thus benefit women too. But if I'm working on water supply anyway, why not think about a multi-purpose pump?

7. Manual rope pumps intended for irrigation end up utilized for domestic uses. Manual rope pumps have been installed all over the world as part of various water programs. In many cases, such as in rural Orissa (in eastern India), the pumps are intended for irrigation. Well, it turned out that in Orissa, the pumps were not used for irrigation. Irrigating a field requires a huge amount of water, and manually pumping that volume of water is time-consuming and exhausting, especially in the heat of the dry season (temperatures can soar to 50 C/120 F). Instead, it turned out that women used the pumps for domestic purposes: drinking, cooking, washing. The women love the rope pumps because they are way easier to use than throwing down and raising a bucket. Plus, children and the elderly can use a rope pump; otherwise they need the women to fetch water for them, since they are not strong enough to raise a heavy bucket of water.

8. Farmers like the color green. This is maybe a silly one, but farmers associate green with agricultural productivity. PRADAN, the grassroots NGO I worked with, told me that whenever possible they use the color green in their products and services, because the farmers respond better to it.

So, what did I do with all of this? I designed a dual-mode rope pump. The pump can be used in motorized mode for high-flow applications such as irrigation and in manual mode for low-flow applications such as domestic uses. The engine is removable, so it can be safely stored at home (farmers expressed concerns about theft) and one engine can be shared by or rented out to several wells, as is done now with the regular centrifugal pumps. Plus the men have no problem allowing women to use the pump in manual mode (the hand crank is removable so it doesn't injure someone when the pump is operated in motorized mode). And yes, I painted the pump green.

Below is a video of the pump in action, being tested by users. I got pretty good feedback from the users, and I have some ideas for future modifications.



Speaking of user testing, I employed human-centered design methodology throughout my research process. Here's a video explaining what I did:



Ok, so maybe this post wasn't a very short summary.

HUGE thank you to PRADAN and Swastik Engineering Works for their support. I could not have done this project without them.

My thesis

I completed and submitted my thesis about two months ago and, finally, earned my master's in mechanical engineering! I very ceremoniously received my diploma in the mail, so it's official.

If you are interested in reading my thesis (or parts of it), please click here. Warning: it's a 15 MB pdf file. I will soon post a summary of my findings, if you don't want to read the thesis but are interested in my work. You can read a bit about the problem I focused on in these posts.

So what's next? Well... I don't know yet. I'm in the process of applying to jobs. If you know of any awesome opportunities in international development/energy/water/all-of-the-above, please let me know!

Back in India again. And behavioral irrigation economics.

I'm back in India. Bet you didn't see that coming! It happened extremely quickly. I literally bought my flights two days before I took off. Highlights of flying Swiss Air through Zurich: free Swiss chocolate and amazing views of the Alps. Highlights of returning to Delhi: catching up with old friends and drinking mango shakes.

Two days ago, my first day back, I met with an agricultural economics researcher. He researches irrigation in Bihar and Gujarat, which is why I was meeting with him. He is interested in the groundwater economy and pumping behaviors. Like me, he finds irrigation fascinating because it lies at the energy-water-food-livelihood nexus.

I learned a lot from this meeting. For one, I learned that economics is really a study of human behavior. I guess that's obvious, but for some reason I never thought of economics that way. This makes econ a lot more interesting than what I had imagined it to be.

As I may or may not have explained before on this blog (I honestly don't remember), in eastern India where pumps are ~85% diesel- or kerosene-run rather than electric, pumps are not installed on a well. Instead, a handful of people own pumps and rent them out to their neighbors on an hourly basis, and pumps are transported on bicycles. These pump owners essentially run an oligopoly. They seem to agree on high rental rates. Interestingly, as more farmers purchase their own pumps and enter the rental business, increased competition has not driven down hourly rates, contrary to what one might expect. This researcher thinks he has learned why: the costs to the pump owner are so high he doesn't have any particular motivation to actually rent out the pump to more customers. He must deliver the pump to the well, which is a pain in the ass. A pump is a pretty heavy thing to strap to your bike. Then, the farmer who is renting the pump may or may not know how to operate the pump, so the owner has to start up the pump for him. Someone has to hang around near the pump to make sure operation is going smoothly and to add more fuel when necessary; sometimes, this someone is the pump owner and not the renter, if the renter is inexperienced with diesel engines. A pump is usually run for several hours at a time, and if the pump owner must babysit the pump for that time, he is losing out on hours that could be spent more productively (in most cases, the pump owner has his own farm to tend to). His time is pretty valuable, so he keeps rental costs high, and often he would rather have that time to do other work than rent out to another customer. Therefore more competition does not reduce prices.

Because of these high operation costs that do not even include fuel cost, according to this researcher, advances in efficiency of the pump would not make much difference to the pumping behaviors of farmers. I'm not taking into consideration all costs involved in operation. Yes, the farmers would spend less on fuel. But the time cost would remain high. Maybe if farmers are getting more water per liter of fuel or per hour, they would be able to irrigate more. But if the farmers want to translate the fuel savings into more hours of pumping, that puts a bigger burden on the pump owner. It is possible that the pump owner would raise hourly rates in response. So even if I make the most efficient pump ever, I might not have any impact on reducing operation costs. (But this doesn't mean a more efficient pump is a bad thing!) In that case, my hope should be that the farmers would get more water for the existing amount of time they irrigate. However, this increased efficiency in operation hardly matters if the pump has a higher capital cost than the cheapest pumps on the market (which, at least at this point, it certainly would). Capital cost reigns supreme over operation costs in financial decision-making. Though all the renters would benefit, the pump owner sees little advantage to his rental business to have a more efficient but more expensive pump--more demand for his pump means more work in renting out the pump, and the rental business is not his only source of income. So why bother spending money on a more expensive pump?

...I really need to learn more about economics.

In addition to enlightening me about pumping behavior and economics, the researcher confirmed something I already suspected: farmers lie about everything when surveyed. (Ok, "lie" might be too strong a word. Stretch the truth, maybe?) But I did not understand the whole picture. I had always thought that I couldn't fully trust people's answers to my survey questions because they were trying to give me the "right" answer. I thought they were trying to come up with the answer they thought I wanted to hear (for example, a woman might lie about keeping her child away from the stove while she's cooking, even if the kid sits right next to her, because she knows I would think the smoke is bad for her child's health). This is true in some cases. However, in many cases, especially when you ask about earnings and expenditures, people exaggerate to make themselves seem poorer. Says the researcher, "They see you, a white woman, or me, a city guy, and they think 'this person works for some NGO and is going to go back to Delhi and write up a report about how we need more subsidies or government assistance.' So they exaggerate how poor they are in order to get more money." Even if a fellow villager is taking the survey, the fact that the survey is taking place at all alters people's answers.

This is a very cynical point of view, and I assume this isn't true for every single person interviewed, but I can believe this does happen sometimes. Probably this behavior stems from a history of NGOs advocating for more government assistance based on field surveys. Like how kids in Kerala constantly ask white tourists for pens because about a decade ago an American group donated supplies to local schools there. (This researcher does not deny that these people are indeed very poor or in need of assistance, by the way.)

So how does this researcher deal with the untrustworthiness of survey responses? "Just add error bars. Uncertainty is part of the fun of social science! See, you want accurate numbers. That's why you're an engineer. You like precision. You don't get the same precision in social science, and you have to be willing to work with that." Well, we use our fair share of error bars in engineering, too. But I guess I see his point.

(To be fair, his research is not all wishy-washy. He gets real numbers when he can: he acquires irrigation data from electricity meters, flow meters, pressure gauges, etc. like an engineer would. But questions of income and costs are a lot trickier to answer in a village where people don't have good records. No receipts, bills, paychecks, etc. Without any paper trail, you have to take people at their word.)

I have been thinking recently about switching into the social sciences (maybe economics or public policy) (this is a topic for another post). But as someone who has training in engineering, will I find the fuzziness frustrating? Or will I find it to be an exciting puzzle to be deciphered, the way this researcher does? I have a lot to think about re: my future.

(As a side note, this researcher got his master's at Princeton and his PhD at Harvard. To those of you interested in studying public policy with a focus in international development, he recommended Harvard over Princeton.)

In other news, I noticed in my blog stats that I get a lot of traffic from a seemingly random blog out there in cyberspace. Apparently, a blogger named Vikram Garg called my blog "the best American in India blog." Thanks for the shoutout, Vikram. Shouting right back at ya! Check out his blog at http://vikramvgarg.wordpress.com/.  

Adivasi Economy and Water Access (or lack thereof)

Pranam dobara, Jharkhand. (Or in English: Hello again, Jharkhand.)

I'm back in Jharkhand conducting a feasibility study for a solar thermal pump. Why a pump? As the tribes of Jharkhand have traditionally been engaged in hunting and gathering, they are relatively new to agriculture and thus have no irrigation infrastructure. Only 5% of the state of Jharkhand is irrigated; the rest rely completely on rainfall (this monsoon season's lack of rain is having serious consequences, which I will discuss later). Why solar thermal? Because diesel is soon to be deregulated and, without subsidies, it will become too expensive for poor farmers to purchase the fuel to operate their diesel pumps (assuming they even have pumps). As it is, legal diesel is not easily available to these communities. The farmers explained to me that they must buy diesel on the black market, and this diesel is often adulterated and thus the pump often does not work properly. Electricity, which is free or close to free for agricultural purposes in India, is either nonexistent or extremely unreliable in these villages. Meanwhile, the capital cost of solar PV pumps is too high. Solar thermal is much less expensive than solar PV, plus the fuel (sunlight, duh!) is free and available, so this could be a good irrigation solution. For some reason, I'm not sure why, it seems no one has attempted to develop a solar thermal pump, other than an NGO in Ethiopia, but they have faced some mechanical issues and their pump is priced too high for Indian farmers. I am also thinking about possibly including a built-in filter or still so that the water that exits the pump outlet is clean, but maybe this is getting too complicated. (I have not made a final decision about what my project will be; next week I will be doing a feasibility study for another, totally unrelated, project in the salt pans of Gujarat.)

I am spending my time in some of Jharkhand's poorest communities: adivasi (tribal) villages in Gumla district. Most villages I have visited belong to the Oraon tribe, who speak a language called Sadri, and the other villages belong to the Khadia and Lohar tribes, who speak their own language as well as Sadri (since Oraon is the dominant tribe in the area, the other tribes have learned their language). Their Hindi is at times difficult for me to understand because (a) it's a different dialect and accent than the Hindi I have learned (it is similar to Bihari Hindi), (b) when they don't know a word in Hindi (after all, it is their second or third language) they substitute a Sadri word, and (c) they don't use the usual English and Urdu loan words that I've gotten used to in Delhi and Shimla--they use the original Sanskritized word.

I have been interviewing farmers about sinchai (irrigation), and, obviously, this involves a lot of questions about khetibari (agriculture) and more generally about their livelihoods. I also stumbled upon a fantastic book in the NGO's office, Mainstreaming the Margins: Water-centric Livelihood Strategies for Revitalizing Tribal Agriculture in Central India* by Sanjiv J. Phansalkar and Shilp Verma. I don't think a more perfect book could exist for the project I'm currently researching. So, let's talk about the adivasi economy. (Note to Jhanvi: you had asked me for some more context to understand the coal cycle wallahs. Here it is.)


*The use of the term "mainstreaming" here does not mean assimilating the tribal communities into mainstream Indian society. The authors emphasize that there is "no inherent conflict" in preserving tribal identity and culture and an approach to tribal development that involves mainstream water technologies and ties to the mainstream markets. Besides, I tend to believe that cultures are dynamic. What culture has really stayed the same throughout the centuries? While there is certainly value in protecting certain aspects of culture, I would argue that it is more important to live a meaningful life free of poverty--not that being free of poverty necessitates sacrificing traditions. Of course it is preferable to both preserve culture and promote development, when possible. Anyway, the Christian missionaries have already altered tribal culture; a huge percentage of adivasis have converted from their animist religions to Christianity. As a reaction to this, Hindu missionaries swooped in and converted many adivasis as well. Fairly few people still practice their traditional tribal religions. Although I'm usually very anti-missionary, I have to admit that they have done some good: in Northeast India, especially Nagaland, Christianity has brought high literacy rates and an end to intertribal warfare. Ok, tangent-rant done.

Phansalkar and Verma explain that the adivasis participate in three economic spheres: (1) forests, (2) agriculture, and (3) migration. It is a very common misunderstanding among the mainstream Indians that all adivasis depend only on forest activities (forest activities basically means gathering "non-timber forest produce (NTFP)" such as wild fruits, tubers, and mushrooms). Each Central Indian adivasi group is different from each other, and while many (if not all) groups have roots in hunting and gathering, their engagement in NTFP today varies widely. The communities I have been visiting in Gumla have largely abandoned that way of life in favor of a settled agrarian lifestyle.

Well, settled to an extent. Phansalkar and Verma emphasize that tribal agriculture is not modernized and thus cannot sustain a community for an entire year. Tribal agriculture is rain-fed and has no irrigation inputs, so during the non-rainy seasons people must migrate to other parts of India to work as laborers. Some bring their families with them, while others send money back to their families.

After reading a bit about migration in this book, I decided to ask the villagers what they do during the rabi (November/December to February) and garmi (March/April to June) seasons if they don't cultivate their land. As expected, many answered that they migrate. I asked to where, and the answer surprised me: to brick industries in Uttar Pradesh and cement industries in Himachal Pradesh.

Wait, did I hear that right? Cement industries in Himachal Pradesh? They couldn't possibly be referring to Nalagarh, where I had worked with IIRD in 2011, could they? (You may or may not remember my two blog posts discussing Nalagarh's cement industry: the first and the follow-up, in which I briefly discussed the migrant workers I had at the time believed to be Bihari.) I asked them, "do you go to Nalagarh?" and now it was their turn to be shocked. "You know Nalagarh?!" "I worked in Nalagarh on village development planning," I told them. Apparently I had been wrong about the migrants in Nalagarh being Bihari; they were Jharkhandi, and from these villages! Who knew these two very remote, very different areas were tied to each other? And what a coincidence that I had worked in both the area that was demanding the migrants and the area supplying the migrants! India continues to astound me with what a small place it is, despite being a country of over a billion people. (Yes, I'm aware I've written that sentence before, possibly more than once. The smallness of India really never ceases to amaze me.)

Phansalkar and Verma argue that this migration is the biggest obstacle to the development of the tribal belt. What good are health and education initiatives if people aren't around to receive the benefits? They claim that the government and the missionaries (who have historically been the only ones helping the adivasis--that's why there are so many Christian adivasis) are attacking the symptoms, not the root cause, of the communities' poverty problems. To lift the adivasis out of their poor living conditions, they must be given assistance to build a more stable livelihood in their home villages, to build a life without migration. Only then will these health and education programs become effective. The key to ending migration? Irrigation that will allow year-round agricultural productivity. The root cause of tribal poverty, then, is poor access to water, according to the authors.

As I stated earlier, right now the vast majority of tribal communities depend solely on rain for their agricultural livelihood. This means most adivasis only grow one season of crops, the kharif (monsoon) crops. Unfortunately (unfortunately is an understatement), this year has seen drought-like conditions (the Government of India is refusing to declare a drought, but the Jharkhand state government is considering it). This is my sixth consecutive monsoon season in South Asia, and I can say it has certainly been the driest I've experienced. It is unbelievable to ride a motorcycle through these villages and pass acres and acres of land covered in grass and weeds from the previous season--at the end of July! Rainfall has been so low that farmers didn't even bother to turn the soil, let alone sow the seeds. Why waste the money on seeds, fertilizer, pesticides, and herbicides if there isn't enough water for the crops to grow? I do see a few tilled plots scattered here and there, but I haven't seen much of the beautiful, healthy florescent green rice paddies that I'm used to seeing in the monsoon.

This lack of rain is terrifying for several reasons: the kharif crops provide these communities with food for the rest of the year. If the kharif fails, then they will not have enough food to eat (and they have very little money to buy food from elsewhere)--people will go hungry. Additionally, people will go thirsty. In every village I have visited, people tell me that their wells dry up by the end of summer (which in India is April to June) and that usually the monsoon rains re-fill the wells. This year, however, the wells have remained dry. These wells provide the only source of drinking water for the entire year, and villagers depend on the rains replenishing these wells during the monsoon. Even after a good monsoon their drinking water supply is limited (this is why there is little to no agricultural activity during the rabi and garmi seasons; the wells do not have enough water for both irrigation and domestic purposes, and the communities consider drinking water more important), so a poor monsoon can be catastrophic. I cannot properly articulate how grave these circumstances really are and how scared I am for these communities for the upcoming year.

What does my pump idea have to do with all this? Well, I hope that by increasing access to groundwater (which is actually quite accessible in Jharkhand, where the water table is high at less than 15 meters), I could help reduce the dependence on rain. Of course the groundwater level itself depends on the rain, so utilizing groundwater wouldn't completely eliminate the problem. However, especially if coupled with groundwater recharge methods, pumps to access groundwater could certainly alleviate some of the issues (worst case scenario, just keep digging deeper until you hit water). The NGO facilitating my visit here has developed earthworks methods that have proven quite successful in aiding groundwater recharge; they have actually seen a rise in the water table where these techniques have been implemented. I hope that, if ultimately I do decide to work on a pump, the implementation would include groundwater recharge earthworks.

Not only would an affordable pump reduce rain dependence during the kharif, it would allow for additional crop seasons. A second (rabi) or even third (garmi) crop season utilizing groundwater irrigation would significantly increase a family's income as well as provide them with a stable year-round livelihood. They would no longer have to migrate for work. And, as I explained earlier, Phansalkar and Verma believe the end to migration is integral to raising these communities out of poverty, because staying put allows them to take advantage of social services.

Bas.

(PS: Sorry there are no photos. The Internet here is waaaay too slow for me to upload any.)