document 525 pre-implementation report chapter: …

Document 525

PRE-IMPLEMENTATION REPORT

CHAPTER: University of Minnesota

COUNTRY: Guatemala

COMMUNITY: Agua Caliente

PROJECT: Expanding Agricultural Opportunities

Prepared By

Nick Bodette

Rebecca Herron

Kim Haglund

Jacob French

Jacob Robole

Lucas Green

Isaac Murphy

Isaac Johnson

Anirudh Srivatsa

Alex Motley

Burke Minahan

12-15-2013

ENGINEERS WITHOUT BORDERS-USA

www.ewb-usa.org

525 - Pre-Implementation Report 12-15-2013

University of Minnesota

Agua Caliente, Guatemala

Expanding Agricultural Opportunities

© 2013 Engineers Without Borders USA. All Rights Reserved of 58

Table of Contents Pre-Implementation Report Part 1 – Administrative Information .................................................. 4

1. Contact Information ............................................................................................................. 4 2. Travel History ...................................................................................................................... 5 3. Travel Team ......................................................................................................................... 6

4. Health and Safety ................................................................................................................. 6 5. Budget .................................................................................................................................. 7 6. Project Discipline ................................................................................................................. 9 7. Project Location ................................................................................................................... 9 8. Project Impact .................................................................................................................... 10

9. Professional Mentor Resume ............................................................................................. 10 Pre-Implementation Report Part 2 – Technical Information ........................................................ 12

1. Executive Summary ........................................................................................................... 12 2. Introduction ........................................................................................................................ 13 3. Program Background ......................................................................................................... 14

3.1 Project Partners ............................................................................................................... 14

3.2 Community Description .................................................................................................. 15 3.3 Community Priorities ...................................................................................................... 15

3.4 Water Sources ................................................................................................................. 16 3.5 Community Relations ..................................................................................................... 16

4. Facility Design ................................................................................................................... 17

4.1 Description of the Proposed Facilities ............................................................................ 17 4.2 Experiment Design Background ..................................................................................... 18

4.3 Design of Experiment ..................................................................................................... 18 4.4 Experimental Methods .................................................................................................... 19

4.5 Results and Discussion ................................................................................................... 20 4.5.1 VALVE #1 ................................................................................................................... 21 4.5.2 VALVE #2 ................................................................................................................... 22

4.5.3 VALVE #3 ................................................................................................................... 23

4.6 Drawings ......................................................................................................................... 25 4.7 Names and Qualifications of Designers .......................................................................... 25 4.8 524 - Draft Final Design Report Comments ................................................................... 25

5. Construction Plan ............................................................................................................... 26 6. Materials List and Cost Estimate ....................................................................................... 28

7. Sustainability...................................................................................................................... 28

7.1 Background ..................................................................................................................... 29

7.2 Operation and Maintenance ............................................................................................ 29 7.3 Education ........................................................................................................................ 30 7.3.1 Workshops and Community Interaction ...................................................................... 30 7.3.2 Introducing New Farmers to the System ..................................................................... 30 7.3.3 Measuring the Increased Output of Existing Pumps.................................................... 30

7.3.4 Introducing New Delivery Height to Community ....................................................... 31 8. Signed Implementation Agreement ................................................................................... 31 9. Site Assessment Activities ................................................................................................. 33






10. Professional Mentor Assessment ................................................................................... 33 10.1 Professional Mentor Name and Role ............................................................................ 33

10.2 Professional Mentor Assessment .................................................................................. 33 10.3 Professional Mentor Affirmation .................................................................................. 34

Appendix A – Existing System Pictures ................................................................................... 35 A.1 Check Valve ................................................................................................................... 35 A.2 Waste Valve ................................................................................................................... 35

A.3 Air Chamber ................................................................................................................... 38 A.4 Exemplary Existing Pumps ............................................................................................ 39

Appendix B – Lab Testing ........................................................................................................ 42 Appendix C – Analysis of Experimental Design ...................................................................... 48

Appendix D – Cited Sources..................................................................................................... 58






Pre-Implementation Report Part 1 – Administrative

Information

1. Contact Information

Name Email Phone Chapter

Project Lead Jacob French [email protected] (314) 629-5100 UMN

Project Lead Rebecca Herron [email protected] (319) 431-5517 UMN

Ram Pump Lead Nick Bodette [email protected] (651) 295-7032 UMN

President Kelly Stifter [email protected] (651) 328-1937 UMN

Mentor #1 Kevin Miller [email protected]

om

(612) 644-1170 MN

Mentor #2

(Travelling)

Kim Haglund kim.haglund@gmail.

com

(651) 308-8147 MN

Mentor #3

(Travelling)

Justin Schnee [email protected]

m

(651) 230-4199 MN

Faculty Advisor Matt Simcik [email protected] (612) 626-6269 UMN

Health and Safety

Officer

Brian Anderson [email protected] (507) 696-3452 UMN

Health and Safety

Officer

Rachel

Orlovsky

[email protected] (262) 498-8413 UMN

Assistant Health

and Safety

Officer

Alex Motley [email protected] (314) 704-9058 UMN

NGO/Community

Contact

Elizabeth

Howland

[email protected]

om

01150249328889 Long

Way

Home






2. Travel History

Dates of

Travel

Assessment or

Implementation

Description of Trip

August

10-23

2011

Assessment Met with representatives from the APROMAC within the

community. Discussed the existing design and use of current

Irrigation Dams and ram pumps, as well as their future goals for

the system. Introduced them to EWB mission and process and

confirmed their desire to form a partnership. Additionally,

performed preliminary analysis of the Irrigation Dam integrity

and surrounding soil quality.

March

10-19

2012

Assessment Met again with the APROMAC and established the foundation

for further assessment with objectives to strengthen water

distribution and storage infrastructure as determined feasible.

The focus of interaction was strictly to discuss possibilities.

Irrigation Dam measurements and benchmarks were taken for

each of five Irrigation Dams. GPS readings were taken over

various locations to create geographical survey of the land.

Community surveys were conducted from members of the

Irrigation Dam association and community members who were as

of yet unaffiliated.

August

17 – 27

2012

Assessment Met again with the APROMAC and signed an MOU for second

assessment. Data collection from the March 2012 assessment was

continued with GPS surveys, land surveys, community surveys,

water testing, soil testing, and hydrologic measurements. Much

work was done on the land surrounding the third Dam due to the

possibility of collapse of the Dam.

August

8-30

2013

Implementation EWB-UMN implemented a dam reinforcement to Dam 3 in Agua

Caliente. Despite minor setbacks, the team managed to work hard

to get most of the concrete poured and left careful instructions for

some hired workers and our NGO coordinator from Long Way

Home to finish the project. The project was entirely completed

approximately 2 months after our departure, and everything was

completed to EWB-UMN’s standards and specifications. We

consider this implementation a success.






3. Travel Team # Name E-mail Phone Chapter Student or

Professional

1 Jacob French [email protected] (314) 629-5100 UMN Student

2 Rebecca Herron [email protected] (319) 431-5517 UMN Student

3 Nick Bodette [email protected] (651) 295-7032 UMN Student

4 Kim Haglund [email protected] (651) 308-8147 MN Professional

5 Jacob Robole [email protected] (612) 229-9205 UMN Student

6 Samantha Meyer [email protected] (763) 607-6920 UMN Student

7 Burke Minahan [email protected] (920) 495-8726 UMN Student

8 Daniel Hoffman [email protected] (715) 803-5666 UMN Student

9 Justin Schnee [email protected] (651) 230-4199 MN Professional

4. Health and Safety The health and safety requirements have already been evaluated and collected in the Health and

Safety Plan. The travel team will refer to this document for any and all safety concerns. Among

other things, this document includes safety procedures and guidelines for in-country security,

worksite safety, and biological/chemical safety.

Security of the travel team is a significant concern. While petty crime has been rated at medium

to high by International SOS, team members can minimize danger by travelling in groups, as

well as always carrying cell phones. Team members should also inform others of where they

plan to go, and an estimated time of return. Finally, danger can be minimized by exercising

reasonable caution.

Main concerns for worksite safety are centered on personal safety. In order to ensure the

wellbeing of any workers on site, personal protection is mandatory. Steel toe work boots and

hefty pants should be worn by any on-site personnel. Gloves must be worn when working with

power tools. Waterproof rain boots should also be worn for any work in the streambed.

The only chemicals the travel team will encounter are chlorine and the metal lubricant,

Molybdenum. To minimize problems, chemical resistant gloves should be worn when

encountering these chemicals. The environment does pose significant biological hazards as well.

In order to minimize risk of disease, the group should avoid drinking the water, and only drink

bottled water from trusted vendors. Team members should also be aware of venomous snakes,

spiders and scorpions. To minimize risk, team members must be aware of their surroundings, as

well as educate themselves on first aid techniques.

Planning, Monitoring, Evaluation and Learning

The travel team has reviewed the 901B – Program Impact Monitoring Report template and

has assigned travel team members to complete this report during the upcoming trip. We






acknowledge that the completed 901B is required with the eventual submittal of the 526 –

Post-Implementation Trip Report. ___Yes ___No

5. Budget Project ID: University of Minnesota Agua Caliente, Guatemala

Type of Trip: Implementation = $3,700

Item Quantity

Unit

Price

Total

Cost

Travel

Airfare 9 round trip flights $850 $7,650

Gas Q1000 to LWH for 1 week $130 $130

Transportation 2 Trips to the airport $80 $160

Misc. To LWH $500 $500

Total $8,440

Travel Logistics

Inoculations Inoculations for 9 people $10 $90

Insurance 9 people $15.75 $142

Total $232

Food and Lodging

Lodging

9 people for 9 nights, 81 man

nights $10 $810

Food and Beverage Dinner @ Feliciano’s 81 times $3.75 $304

Misc. Gratuity $20 $20

Total $1,134

Labor

In-country Logistical

Support

Internet access at LWH for 1

week $25 $25

Local Skilled Labor Half days work for welding $200 $100

Misc. Translator for the week $1,000 $1,000

Total $1,125

EWB-USA

Program QA/QC Implementation $3,700 $3,700

Total $3,700

Materials

4in Flange 20 $12 $240

4in Elbow 5 $15 $75

4in Pipe 3ft $5 $15

Bolt 85 $0.80 $68






Washer 160 $0.33 $53

Nut 100 $0.40 $40

Rubber Gasket 10 $5 $50

Threaded Rod 7 ft. $1.50 $11

Extra Rubber 2ft^2 $5 $10

Metal Bushing 5 $1 $5

Metal Weights 30 $1 $30

Total $596

Grand Total $15,227

EWB-USA Headquarters use:

Indirect Costs

EWB-USA

Program Infrastructure (2) See Below $0

Sub-Total $0

TRIP GRAND TOTAL (Does not include Non-Budget Items)

$0

Program Infrastructure (EWB-USA Headquarters accounting, administration and

fundraising)

Assessment = $500

Implementation = $1,200

Monitoring = $350

Non-Budget Items:

Additional Contributions to Project Costs

Community

Labor 2 Laborers for Ram Pump Construction

Materials All materials required to assemble and

improve valves and pumps

Logistics None

Cash None

Other Community members to attend ram pump

workshops

Community Sub-Total

EWB-USA Professional Service In-Kind

Professional Service Hours 144

Hours converted to $ (1 hour = $100) $14,400

Professional Service In-Kind Sub-Total $14,400






TRIP GRAND TOTAL (Includes Non-Budget

Items) $0

Chapter Revenue

Funds Raised for Project by Source

Actual Raised

to Date

Source and Amount (Expand as Needed)

Engineering Societies $1000

Corporations $0

University $3000

Rotary $0

Grants - Government $0

Grants - Foundation/Trusts $4000(project)

Grants - EWB-USA program $0

Other Nonprofits $0

Individuals $2400

Special Events $247.26

YEC $4000(project)

Misc. $0

EWB-USA Program QA/QC Subsidy (3) See below $ 0

Total $13,647.26

Remaining Funds Needed $1579.74

Program QA/QC & Infrastructure Subsidy:

Assessment = $1450

Implementation = $3,800

Monitoring = $950

6. Project Discipline Water Pumps and distribution.

7. Project Location

Agua Caliente Community Center:

Longitude: 14°48’18.76” Latitude: 90°51’71.23”






8. Project Impact Number of persons directly affected: 150

Number of persons indirectly affected: 400

9. Professional Mentor Resume Kim Haglund

Education:

University of Minnesota, Twin Cities

Bachelor of Biomedical Engineering, 2007

Minor: Spanish Studies

Experience:

R&D Engineer

Vision-Ease Lens, Ramsey, MN

February 2008 – Present

- Responsible for development, improvement, and coordination of tests on ophthalmic

lenses and raw materials

- Oversee mechanical, analytical, optical, performance, and reliability testing on such

technologies as injection molding, cast molding, dip coating, vacuum coating,

photochromic dyes, and polarized film

- Work closely with Quality organizations in the U.S., Thailand, and Indonesia to ensure

consistency in test methods and results

- Supervise a team of one scientist and four technicians

- Manage the safety and organization of all research laboratories

R&D Engineer, Sustaining Engineering

Boston Scientific Corporation, Maple Grove, MN

June 2007 – November 2007

- Supported commercialized products in stents and coatings

- Collaborated on various projects with Operations, Quality, Marketing, and Regulatory

Affairs

Research Assistant, Department of Chemical Engineering and Materials Science

University of Minnesota, Minneapolis, MN

February 2004 – May 2007

- Conducted research on fuel cells, including ink-jet and laser-printed electrodes

- Performed electrode position testing of Platinum and Platinum/Ruthenium anodic films

- Managed and maintained the “Dry Room”, a low-humidity laboratory






R&D Intern

Boston Scientific Corporation, Maple Grove, MN

May 2006 – August 2006

- Conducted extensive literature research on PLGA-based coating for drug-eluting stent

- Prepared and tested polymer samples for dry and wet adhesive strength






Pre-Implementation Report Part 2 – Technical Information

1. Executive Summary Expanding Agricultural Opportunities is a mechanical project proposed by the University of

Minnesota Engineers Without Borders Chapter (EWB-USA UMN). An implementation is

planned, during which EWB-UMN will hold several workshops with the community, explaining

the improved design and maintenance requirements for a series of existing hydraulic rams.

EWB-USA UMN has carried out extensive testing on a full scale model of the ram pumps found

in Agua Caliente with EWB-USA UMN designed valves, as well as a replica of a valve found in

Agua Caliente. EWB-USA UMN is requesting TAC approval for the experimental methods,

design, and results of the chapters ram pump study in order to share the results with the

community.

The overall goal of this project is to improve APROMAC’s entire agricultural water distribution

system in Agua Caliente. EWB-UMN recently completed a project to reinforce the biggest and

most exposed of the five dams in the Agua Caliente system, dam 3. The remaining dams were

either structurally sound, or not feasible to repair due to location or inherent design flaws. After

the integrity of the system was secured, the logical next step in the project was to improve the

efficiency and power of the ram pumps, thus improving the usefulness of the system to the

community, while potentially allowing more people to participate in the APROMAC co-op

because of the increased capability and range of the rams.

Our NGO, Long Way Home, has a long standing relationship with Agua Caliente, and has

helped us with communication with the community. Agua Caliente is a small farming

community in central Guatemala, which is home to APROMAC, an agricultural co-op focused

on blackberry production, which owns the agricultural water distribution system. APROMAC’s

system was originally designed by the community of Agua Caliente, which consists of a series of

five dams that vary in size, design, and structural stability. Installed on each of the dams are

several ram pumps that pump water up to agricultural fields at elevations above the stream. This

allows the community to grow crops in the dry season, increasing their yield and profit. This

allows their children to go to school longer improves the nutrition of the town, and improves the

economy of the community and surrounding area because the farmers with ram pumps hire

outside workers to tend their fields in the dry season. The community reached out to our group in

order to reinforce and improve the system, which is quickly falling apart because of lack of

knowledge and experience. The dams were replicated based on a local Guatemalan engineer’s

design, and the rams were developed by the community based on an encyclopedia entry on

Persian pumps. Because everything was originally built by the community, we know they have

invested significantly in the system already. EWB-USA UMN has drafted a Memorandum of

Understanding (MOU) between EWB-UMN and APROMAC, which can be found in section 9

of this document, and is currently working with Long Way Home and the community to get the






document signed. We keep in contact with the community via the proposed monthly calls

facilitated through Long Way Home.

In August 2013 EWB-UMN successfully completed an implementation to reinforce dam 3 on the

Agua Caliente system. Two assessment trips were taken prior to the dam implementation.

Additionally, EWB-UMN has been very active in the greater Agua Caliente area, including the

construction of a large scale rainwater harvesting system in the nearby community of

Simajahuleu. All of these previous trips have been in conjunction with Long Way Home.

Due to the complex mechanics and lack of accurate mathematical models of ram pumps, our

chapter opted to make a model ram pump to test under various stroke lengths and valve weights.

Minitab was used to analyze the experimental results. This full analysis can found under section

5.3.

Creo was used to make a 3D model of the ram pumps found in Agua Caliente, based on detailed

measurements taken during previous assessments and the last implementation. The complete

drawing set can be found in the PDF attached with this document.

Due to the heavy design emphasis of this project, it was decided that the best way to disseminate

the new design information was through a series of workshops with the community, during

which new valves will be built and assembled with an existing ram body with members of

APROMAC. This process will give the each member of APROMAC access to the new design

and allow them to make improvements on their individual pump as they see fit. This will also

allow prospective new members to see how to design the pumps and use them if they decide to

join the system.

2. Introduction Since August 2011, EWB-USA UMN has been working in the community of Agua Caliente to

improve a deteriorating agricultural water system, which has been functioning since 1990. The

system is operated by a local farming cooperative called APROMAC and consists of 42 ram

pumps and five dams, as well as a handful of fields which contain water storage tanks and micro-

irrigation pipes. EWB-USA UMN has been working closely with APROMAC and our NGO,

Long Way Home, to determine the best way to improve the system. In March and August 2012

it was determined that Dam 3 was in critical need of reinforcement or it would soon fail. This

was EWB-UMN’s third implementation in the area, and first implementation in the

community. For our next implementation we hope to have a more widespread impact on the

system because the last implementation was so focused on helping a specific cohort of the

community. For this reason, improvement of the ram pumps was chosen as EWB-USA UMN’s

next implementation in Guatemala because of their widespread impact on the existing system

and the community.

The current system allows farmers to grow crops during the normally unproductive dry

season. Before, many farmers had to travel to the coast or other regions of the country to find






work during the dry season. Now they are able to find work within the community year-

round. This has allowed what once was subsistence farming to turn into a commercial

venture. Currently the APROMAC has Q7000 per pump and Q200, 000 per dam invested into

the system, and is therefore the greatest beneficiary to any improvements done on it. However,

expanding the system will extend benefits to other members of the community who were not

able to use the system before. There are several people in the community who would like to

utilize the system but are unable because their crops are too far from the water source for the

current pump capacity. By increasing the output specs of the ram pump design, we hope to be

able to allow farmers to join the system that were not able to at the time of its creation. Adding

our changes to the system we estimate will cost approximately Q245 per pump, a very small

amount when compared to the cost of a new pump.

3. Program Background

The University of Minnesota has worked in the Comalapa/San Juan region since 2006. During

this time two implementation projects have occurred (a spring box and pump in Chimiya, and a

rain water harvesting system (RWHS) in Simajhuleu). In August 2011 we visited a new

community, Agua Caliente, to meet with community members and gather preliminary

assessment data.

3.1 Project Partners

Long Way Home (LWH). “LWH is a non-profit organization which uses sustainable design and

materials to construct a self-sufficient school that promote education, employment and

environmental stewardship”, and is based out of Comalapa, Guatemala. EWB-USA UMN has

been partnered with LWH for several years and has worked with them to install a spring box and

pump in Chimiya, a rainwater harvesting system in Simajuleiu, and most recently a dam repair in

Agua Caliente. They have continued to work with us as translators, planners, and advisors

throughout the Agua Caliente project.

Asociacion de Desarrollo Integral de Productores de Mora Agua Caliente (APROMAC). The

APROMAC is a farming cooperative that has funded the irrigation dam and ram pump system

that is located on the stream leading from Patzá. They are forward-thinking and community-

minded, and have shown a lot of interest in working with EWB-USA UMN. Although they

already have an irrigation water system, much of it was designed without the aid of engineers

and is quickly falling apart, thereby causing erosion and becoming a danger to their investment

and lifestyle.

COCODE. The COCODE is the official local leadership board which oversees the community,

including the chlorinated drinking water system of Agua Caliente. Some members of the

COCODE are also members of the APROMAC, and both groups seem interested in working

with EWB-USA UMN to improve their irrigation system. Though EWB-USA UMN is not

working in direct relationship with the COCODE authorities, the president of the COCODE is a

member of the APROMAC and we continue to listen to COCODE members for suggestions or






community-wide goals.

EWB-USA MN Professional Group. Once every three months during 2013, the EWB-USA MN

Professional group will attend a EWB-USA UMN project meeting and will offer assistance to

our designs and methodology. The EWB-USA MN Professional group will also provide

mentorship to our group and will act as a reference for structural designs. EWB-USA UMN has

a very strong relationship with this group.

3.2 Community Description

Agua Caliente is composed of 200 families and is located on the Rio Agua Caliente. The total

population of the community is estimated to be around 1400 people, which averages around

seven people per family. There are two schools, a basic school and a middle school, but no

health clinic or market. For trade or for serious medical problems they go to the neighboring

villages of Simajahuleu or Poaquil, and if necessary, to the city of Comalapa. Their political

system is typical of others in the area and is composed of an Alcaldes and a COCODE.

Twenty years ago the community used to sell wood and carbon products but, after noticing the

rate of deforestation, decided to look for a more sustainable source of income. In response, 70

families organized to create a farming cooperative called the APROMAC, which also helped to

improve the health of their crops and protect them from larger buyers by selling crops as a

group. Since 1990, the APROMAC has successfully installed five irrigation dams, with several

ram pumps on each dam, in the stream that leads from Patzá. This enables them to irrigate their

crops in the dry season and, as an added benefit, creates several agricultural jobs for the

community as a whole. They have asked EWB-USA UMN to partner with them to protect the

integrity of the irrigation dams, to help them to improve and expand the ram pumps, and to

possibly add water storage for micro-irrigation in the future.

EWB-USA UMN has decided to define the APROMAC as our community and work with them

apart from the rest of the community of Agua Caliente for several reasons. First, the members of

the APROMAC have invested the most time and money into the irrigation system and

collectively own all of the dams and most of the ram pumps. Second, the APROMAC has

shown a lot of interest in working with EWB-USA UMN to improve the quality of the system

and is willing to provide labor and materials for an implementation. Third, the APROMAC

desires to expand the system to include as many families as possible, thus helping the entire

community of Agua Caliente.

3.3 Community Priorities

The APROMAC has invested a substantial amount of time and money into the current irrigation

system and their top priority is to protect their investment from being degraded or destroyed,

such as by a heavy rainfall. Once this prerequisite is met, the APROMAC has expressed interest

in the following projects, listed in descending order from most important to least important

(approximately, according to APROMAC leaders during the first two assessment trips):

1. Ensure that the dams will not be destroyed in a high rainfall event.






2. Replace the current system of sprinklers with water storage tanks and micro-

irrigation pipes. This will allow farmers to use less water, which will further allow

them to water more fields in a day than they are currently capable of or water multiple

fields at once.

3. Add more pumps, and therefore families, to the dams. This will allow more fields

to be watered.

4. Improve ram pump efficiency and pumping power. This will allow more fields to

be reached and will shorten the time required to water a field or fill a water storage

tank.

5. Replace the current black rubber hosing that connects the ram pumps to the

sprinklers with PVC or another, stronger material. The current range of some pumps

is limited because the rubber hoses break under pressure, and by switching to PVC

more fields will gain access to water.

6. Install a tilapia pond. This will provide the community with an alternative source

of meat.

3.4 Water Sources

There are at least two sources of water that the community uses. One is a hot spring, Patzá,

which flows into the Rio Agua Caliente, and the other comes from a neighboring stream.

Potable drinking water for the community is provided from the neighboring stream and is

chlorinated, and based upon community feedback the water is perfectly safe to drink. However,

there are some community members, especially the elderly, who do not like the taste of the

chlorination and continue to boil their water before drinking. The stream that runs from Patzá is

used primarily as agricultural water, although there have been reports of people swimming,

doing laundry, and letting animals drink from the stream.

3.5 Community Relations

During the repair and monitoring trip for the rainwater harvesting system in Simajhuleu, the

community of Agua Caliente learned about our presence in the area and approached Long Way

Home with interest in working with our group. EWB-USA UMN took a day trip to Agua

Caliente during this trip to see the agricultural system for the first time. We took our first

assessment full assessment trip in March 2012 to examine the system and meet with leaders from

the APROMAC and Agua Caliente. Our primary tasks included: a GPS survey of the

surrounding area; measurement of ram pump efficiency, flow rate, and power; measurement and

analysis of the structure of the dams and their stability; measurement of the spring safe yield

leading from the spring; estimate of the agricultural demand of the community; survey of the

types and acreage of crops that are fed by the irrigation system; testing and qualification of water

and soil at the river and in the fields; conduction of approximately 20 community surveys to gain

community opinion of the APROMAC and to determine the needs of the community.

From the first assessment trip it became clear that two of the dams, hereafter referred to as Dam

2 and Dam 3, were in danger of collapse. To gather more data we took a second assessment trip

in August 2012. Our primary tasks included: a more extensive GPS survey of the surrounding

area; a land survey of the area surrounding the dams, especially Dam 3; conduction of another 25






community surveys, bringing the total number of community surveys to 45; further water and

soil testing and qualification, especially around Dam 3; analysis of the hydrologic structure

surrounding the dams.

Currently, EWB-USA UMN is working towards our second implementation in Agua Caliente.

The following pages outline the methodology that EWB-USA UMN used to determine the nature

of our second implementation.

4. Facility Design

4.1 Description of the Proposed Facilities

This proposal is for increasing the efficiency of the ram pumps in Agua Caliente, Guatemala.

APROMAC, a blackberry co-op in Agua Caliente has created a system of five dams, each

supplying head to ram pumps used to irrigate fields during the dry season. EWB-USA UMN has

already completed one implementation on this system to fortify dam 3, and is now focusing on

increasing the supply distance of APROMAC’s pumps. Improved pumps will supply more water

to the fields currently irrigated by the system and allow more fields to be irrigated, which will

open the potential for additional families to join APROMAC.

The current APROMAC pump design consists of a drive pipe, air chamber, impulse valve, check

valve, and ram body. Water flows from the higher elevation behind the dam, through the drive

pipe and around the impulse valve. As the water reaches a certain velocity, the drag force on the

impulse valve becomes greater than the restoring force and the impulse valve shuts. The sudden

pressure spike caused by the impulse valve shutting causes the check valve between the ram

body and the air chamber to open, and water is pushed through the air chamber and into a supply

hose up to a certain delivery height. Two very crucial parameters of the impulse valve are the

restoring force and the distance the valve is allowed to open, called the stroke. The current ram

pumps used by APROMAC use boot rubber as the restorative force, which is unreliable. The

rubber wears out quickly and is not adjustable as the stroke and restorative forces can’t be

changed. The check valve is the other moving part of the pump. The current check value used by

APROMAC uses the water’s upward force to bend a very thick piece of tire rubber allowing

water to flow around it.

The impulse valve is the easiest piece of the current design to modify from both a monetary and

technical standpoint. Designing a new impulse valve will keep the upgrade cost per pump low

making the pump more affordable and, thus, giving more families the opportunity to upgrade or

join APROMAC. Experimental results showed that a tunable impulse valve with a three inch

impulse valve stopper would allow each pump to be optimized for its’ specific drive pipe length,

head and delivery elevation, thus, allowing each family to get the optimum performance out of

their existing irrigation system. Several educational workshops will be held with existing and

potential APROMAC members to demonstrate how to construct, tune and maintain the new

valve design. The estimated per cost per new impulse valve is thirty-five U.S. dollars, which is






just a small fraction of the total cost of a ram pump. Therefore, we hope that most members of

the community will find this design change very cost-effective.

As well as holding educational workshops for the proper construction and tuning of the newly

designed impulse valve, we plan on talking to the community about the potential for

implementing a drip irrigation system. If the community expressed interest in a drip irrigation

system, it would be a community-led implementation, and we would only serve to ask Toro to

donate the necessary materials. We also plan on asking the community if they have any interest

in organic farming, and potentially adapting to organic farming practices.

4.2 Experiment Design Background

Since hydraulic ram pumps are hard to analyze and predict with theory, we had to take a more

qualitative approach to redesigning the impulse valve. Also, since we could not exactly replicate

the current system in the community, and because the system in the community varies so much

from pump to pump, we decided to compare the flow rate between a replica of the valve

currently used in the community, and two valves EWB-USA UMN designed instead of

comparing the results to data from the ram pumps operating in the community. Therefore, we

had to design and build an impulse valve that would closely match the community’s current

design to test and compare with the other impulse valve designs EWB-USA UMN created.

Because we had to approach this in a qualitative manner, we decided to do some research on

what types of valves have been used in other hydraulic ram models. The two considerations we

took into account while designing the impulse valve were the ability to tune the valve and the

simplicity and ruggedness of design. We wanted a valve that is easy to maintain and last for a

long time, and that is easy to properly tune. One of the major problems with the current impulse

valve design in the community is that there is no way to properly tune the stroke length. The

spring force is tunable, but the rubber springs wear out very quickly and need to be replaced

every two weeks.

The two most common impulse valve designs use a spring as the restoring force, and weights as

the restorative force. For simplicity of construction and ease of maintenance and tuning, we

decided to go with an impulse valve design that uses weight as the restoring force. Because of

that, we had to run the impulse valve slider rod vertically, so we also introduced a ninety degree

elbow to the current design. This is a very common impulse valve design, and has been utilized

in many commercial ram pumps. Next, we were tasked with designing this impulse valve such

that the design facilitates constructing it properly, making sure that the slider rod runs coincident

with the center of the impulse valve hole, and that the impulse valve stopper seats square on the

underside of the impulse valve plate. We came up with a simple and efficient design laid out in

the next section.

4.3 Design of Experiment

The performance of hydraulic ram pumps is a function of several variables making it very

complex to predict analytically. Therefore, we decided early on in this project that we needed to






approach this project empirically, as there was no way we could analyze the existing pump and

decide what to change to make it perform better. As discussed above in section 4.2, we chose

two new types of impulse valves, based on commercial designs, to test against their current

impulse valve design.

We were fortunate enough to obtain lab space at the St. Anthony Falls Laboratory (SAFL) at the

University of Minnesota. Working in this versatile hydrology and fluid mechanics research

laboratory allowed us to take water indirectly from the Mississippi River, and then send it back

into the river. Other lab spaces we were considering would have required us to design and build a

water return system to conserve water, but at SAFL all the water is simply returned to the

Mississippi River.

There are several factors that affect the flow rate of a ram pump with a given impulse valve

design. Such factors include the height of the supplied water, the height of the delivered water,

and the impulse valve tune settings, which include the stroke and the added weight of the

impulse valve. To determine which impulse valve pumps the highest flow rate of delivered

water, we held the supplied water and the delivered water height constant throughout our testing,

and varied the tune settings. We used a statistical analysis to determine the best combination of

the stroke and weight for maximum output flow rate for a given impulse valve. The supplied

water and delivered water height were not chosen to represent the conditions in the community;

the supplied water height was constrained by our lab space, and the effective delivered water

height was chosen as an appropriate height to which they could potentially be pumping. We were

not concerned with these values as long as they were in an appropriate range. The purpose of this

experiment is to compare the results within the experiment, and not with the results of the pumps

in the community.

The only metric we are using for analysis outcomes is the delivered flow rate. The power and the

efficiency of a ram pump are two common parameters to examine, but we decided that are only

concern was producing the most amount of flow at the output. It has been shown in past

experiments that the maximum flow rate does not correlate to the maximum efficiency [2].

Therefore, due to the fact that the efficiency and the power are not necessary metrics in our

analysis, and the complications involved in making the necessary measurements, we decided to

only measure the output flow rate. We did this by inferring the flow rate from the weight of

water pumped in one minute.

4.4 Experimental Methods

A detailed experimental procedure is outlined below. We have tested and analyzed the two new

impulse valve designs and the replica of the current impulse valve design being used in the

community. The experimental procedure is very similar between the different valves, with the

only difference being that the two new valves use weights and the current design uses a rubber

spring.

1. Set the correct restoring force for the trial we are running.






a) For the weights, weigh the combinations of the added weights and add them to the top of

the slider rod.

b) For the spring, the restoring force is defined as the force in the spring when the valve

stopper was approximately 1/8” open. Attach a force gauge connected to the slider rod and hold

at the position defined above. Clamp the two sides of the rubber spring onto the back side of the

flange and measure the force. If the tension needs to be adjusted, make the rubber tighter or

looser and then clamp it down and measure again.

2. Set the correct stroke for the trial to be run. This was done by adjusting a nut on the slider rod

that limited the motion of the slider rod and ultimately the valve stopper.

3. Add the correct size snifter plug.

4. Cover the impulse valve and snifter with splash covers.

5. Slowing turn on supply water flow until water flows out the impulse valve gently. Once water

starts to flow out impulse valve, turn water supply on all the way, such that the valve is wide

open.

6. Close throttle valve to allow air chamber to build up pressure. At the beginning, manually

pump the impulse valve if it is not running on its own.

7. Once the pressure gauge shows 60 psi, start to slowly crack open the throttle valve allowing a

small stream of water to flow out. Fine tune the throttle valve such that the maximum pressure

spike hits 60 psi.

8. Once the pump reaches steady operation, allow the delivered water to flow into a five gallon

for one minute, recorded with a stop watch. During that minute, record the number of cycles as

well.

9. After collecting the delivered water for one minute, turn the supply valve off all the way and

close the throttle valve all the way before the pressure drops too much. Observe the pressure

gauge to make sure the pressure is maintained in the air chamber. If the pressure is not

maintained in the air chamber, that would be a possible symptom of the check valve

deteriorating.

10. Open the throttle valve to release the pressure stored in the air chamber.

11. Measure the weight of the water and the bucket together, and then of the bucket all by itself

to determine the weight of the water alone.

4.5 Results and Discussion

The experiment was analyzed using Minitab software in order to fit a model to the data. All

terms (linear, squared, and interaction) were included in each analysis. Any term with a p-value






less than 0.050 was considered to be a significant factor. The delivered water weight was used

as the response.

4.5.1 VALVE #1

The stroke, valve weight, and both of their squared terms were found to be significant. The

interaction term was found to be insignificant. An R-squared value of 86.21% shows that the

model fits the data well. However, two trials varied significantly from the model. Upon

examination of the data, one trial had a very low response that can likely be attributed to a small

stroke length. Since the stroke length in that trial is considered to be unrealistic and far out of the

actual operating range, the data was re-analyzed excluding this trial. Detailed results from this

analysis can be seen in Appendix C.

Again, the stroke, valve weight, and both of their squared terms were found to be significant

while the interaction term was insignificant. The R-squared value improved to 97.33% which

means the model still fits the data very well. One trial varied significantly from the model but

was kept in the analysis due to the excellent fit. The residual plots show that the data is normal

and that there is no evidence of nonconstant variance or error correlation. A response

optimization analysis produced expected optimal values of a 1.78-in stroke length and a 2.88-lb

valve weight. Using these values, the predicted water delivery is 27.2lb. Detailed results from

this analysis can be seen in Appendix C.

Raw Data for Valve #1.

P-Values for all terms.

Trial Stroke (in)

Valve Weight

(lb)

Delivered Water

Weight (lb)

1 2.50 1.00 15.00

2 0.50 1.00 11.00

3 0.50 4.05 18.50

4 2.50 4.05 20.90

5 0.19 2.46 3.70

6 2.90 2.52 23.96

7 1.50 0.38 12.22

8 1.50 4.62 19.32

9 1.50 2.54 25.48

10 1.50 2.54 26.28

11 1.50 2.54 26.85

12 1.50 2.54 27.50

13 1.50 2.54 26.75






Critical Values.

4.5.2 VALVE #2

The stroke and valve weight were found to be significant. Both squared terms and the

interaction term were found to be insignificant. An R-squared value of 80.00% shows that the

model fits the data fairly well, and no trials varied significantly from the model. The residual

plots show that the data is normal and that there is no evidence of non-constant variance or error

correlation. The relationship between the stroke and the valve weight is linear, and therefore

optimal values for these factors could not be found. A linear relationship across all ranges is not

expected; it is more likely that the output will level off or begin to decrease outside of the range

that was tested. If the relationship were linear, a significantly higher stroke length and valve

weight would be required in order to achieve an output comparable to that of valve #1, and these

conditions would be impractical for this design. If the output levels off or begins to decrease, it

is unlikely that the maximum possible output is comparable to that of valve #1. Therefore, no

further testing was done on this valve. Detailed results from this analysis can be seen in

Appendix C.


Term P-value

Stroke 0.013

Valve Weight 0.001

Stroke*Stroke 0.002

Valve Weight*Valve Weight 0.000

Stroke*Valve Weight 0.564

R-squared Value 97.33%

Optimal Stroke Length 1.78 in

Optimal Valve Weight 2.88 lb

Expected Delivered Water 27.2 lb






P-values for all terms.

4.5.3 VALVE #3

The original designed experiment for this valve called for a spring tension of approximately 50

lbs to be tested. During testing, while a spring tension of up to 52 lbs could be achieved, the

pump did not run above 48 lbs. The R-squared value for the model was 38.93%; however, one

trial varied significantly from the model. The data was re-analyzed excluding this trial. The

remaining data is normal, although weakly so (p-value of 0.055). The R-squared value improved

to 73.81%. While this indicates that the model is fair, the model does not appear to be reliable

above 45 lbs of spring tension. Even if the model were in fact accurate, it predicts that a spring

tension above 45 lbs would be required to reach an output comparable to valve #1. Since it is

already known that the valve will not operate at that level, no further testing was done on this

valve. Detailed results from this analysis can be seen in Appendix C.


Trial Stroke (in)

Valve Weight

(lb)

Delivered Water

Weight (lb)

1 1.25 3.02 5.64

2 1.25 3.02 6.50

3 1.25 3.02 6.20

4 1.25 3.02 7.72

5 1.25 3.02 8.16

6 1.25 3.02 8.52

7 0.50 1.52 4.40

8 0.50 4.54 6.70

9 2.00 4.54 9.56

10 2.00 1.52 6.18

11 1.25 0.88 4.68

12 1.25 5.12 11.20

13 0.19 2.99 4.14

14 2.31 2.99 10.58

Term P-value

Stroke 0.005

Valve Weight 0.003

Stroke*Stroke 0.819

Valve Weight*Valve Weight 0.739

Stroke*Valve Weight 0.699






P-values for all terms.

4.5.4 Discussion

As discussed above, we expect that our new design (impulse valve three) will increase the

pumping capacity of the ram pumps in the community. Along will pumping more water, the new

valve will be much easier to tune properly. We found in our experimental runs of the ram pump

that it was very easy to add more weight to the new valve design, but it was very difficult to

increase the tension in the rubber consistently, especially without a force gauge. We expect that

our new valve design will make the tuning process much simpler and quicker. Another problem

with the current design is that the horizontal slider rod can easily bind in the bearing if the rubber

spring is pulling more on one side than the other when the valve is fully shut. This event

completely stopped the pump while we were testing it. This can be avoided by lubricating the

slider rod or make the bearing a looser fit. However, there will be more wear on the slider rod

and the bearing, and it would require regular lubrication. We expect that our new design will

decrease the rate of wear on the slider rod and bearing. Also, in our experimental trials the slider

rod never got caught in the bearing, which could have stopped the pump. Qualitatively, our new

impulse valve is a more robust design which facilitates the tuning process. Quantitatively, we

Trial Stroke (in)

Spring Tension

(lb)

Delivered Water

Weight (lb)

1 1.00 30.00 8.58

2 1.00 48.00 8.48

3 2.50 45.00 10.96

4 2.50 33.00 11.92

5 1.80 28.00 10.00

6 1.80 52.00 DID NOT RUN

7 2.80 40.00 9.92

8 1.80 38.00 5.78

9 1.80 38.00 9.36

10 1.80 38.00 9.78

11 1.80 38.00 9.32

12 1.80 38.00 9.58

13 1.80 38.00 10.64

14 0.70 37.00 15.40

15 1.80 50.00 DID NOT RUN

Term P-value

Stroke 0.061

Spring Tension 0.079

Stroke*Stroke 0.011

Spring Tension*Spring Tension 0.082

Stroke*Spring Tension 0.030






expect to potentially double the flow rate of the pumps in the community with our new impulse

valve design.

4.6 Drawings

See attached PDF for ram pump CAD designs.

4.7 Names and Qualifications of Designers

Name Student or

Professional

Qualifications Work Done

Nick Bodette Student Senior mechanical

engineering student

Design and build of ram model,

leader of ram pump group,

principal lab tech.

Kim Haglund Professional R&D Engineer Design of experiment, data

analysis, mentor

Becca Herron Student Junior Computer

engineering student,

project co-lead

Lead writer on EWB documents,

started ram pump project, helped

with lab testing

Tom Johnson Professional Professional

mechanical engineer

Designed impulse valve

Isaac Murphy Student Senior Mechanical

Engineering student

Designed the CAD drawings,

helped with pump testing

4.8 524 - Draft Final Design Report Comments

Project Design/Technical Description Comments:

No. Pg.

No.

EWB-USA PM Comment Chapter Response

1 You are building a demonstration pump at the

site for farmers to duplicate – is this correct?

This is correct.

2 All material (bar none) must come from

locally. All tools and equipment used to

assemble must be available locally.

Everything is available locally.

3 18 I understand that it will be the responsibility

of each individual who wants a pump to pay

for a construct his own pump – is this correct.

It seems like $625 is a high cost for a pump –

can anyone in the community afford this?

$500 is for pipe already there. The valve is

just $35 - $50.

4 What is the difference between this pump and

the existing pump in terms of water delivered?

What is the improvement?

Same head – higher flowrate. Also higher

head available.

5 Much of existing pipe for each pump can stay

the same.






Next Steps Comments:

No. EWB-USA PM Comment Chapter Response

1 When do you plan on implementing? March? Yes.

2 When do you plan on submitting your 525 pre-

implementation report? December deadline?

Yes.

3 I will not call you after you submit your 525 pre-

implementation report. You will just be scheduled

for a TAC meeting.

Discussed.

5. Construction Plan Our methods for construction are based primarily around allowing the community to observe the

entire process: design of the valve, assembly of the valve, assembly with the pump, and tuning

the valve on the pump. In order to do this well EWB-UMN is designing a series of workshops

that involve machining from raw materials, demonstrating the valve on land, demonstrating

adding the valves to the existing pumps, and showing community members how to tune the

pumps. The existing design does not allow for any tuning whatsoever, but EWB-UMN’s

improved valve design allows for tuning based on stroke length, stroke force, and snifter size.

This allows for maximum pumping capacity for a variety of different needs whether that is high

elevation pumping or distance pumping. Below is a basic schedule for the workshops and

meetings with the community.






Time March

14th

March

15th

March

16th

March

17th

March

18th

March

19th

March

20th

March

21st

March

22nd

March

23rd

7:30

AM

Travel

Day Intro

meeting

with

APROM

AC

Build

impulse

and

check

valves

for ram

pumps

Meet

with

communi

ty

members

intereste

d in

joining

system

Demonst

rate

valve

design on

land and

answer

questions

about

assembly

and

materials

.

Replace

valves in

streambe

d

Demonst

rate in

streambe

d

performa

nce to

communi

ty

members

Meet

with

communi

ty to

discuss

future

plans/pro

jects

Continge

ncy Day

Travel

Day

8:30

AM

9:30

AM

10:30

AM

Meet

with

APROM

AC to

discuss

where to

add

members

to system

11:30

AM

Build

impulse

and

check

valves

for ram

pumps

Test the

valve to

make

sure it

works in

country

Educate

possible

recipient

s of

valve

improve

ment

Tune ram

pumps in

streambe

d with

new

valves

12:30

PM

1:30

PM

2:30

PM

Demonst

rate in

streambe

d

performa

nce to

communi

ty

members

3:30

PM

4:30

PM






6. Materials List and Cost Estimate

Materials

4in Flange 20 $12 $240

4in Elbow 5 $15 $75

4in Pipe 3ft $5 $15

Bolt 85 $0.80 $68

Washer 160 $0.33 $53

Nut 100 $0.40 $40

Rubber Gasket 10 $5 $50

Threaded Rod 7 ft. $1.50 $11

Extra Rubber 2ft^2 $5 $10

Metal Bushing 5 $1 $5

Metal Weights 30 $1 $30

Total $596

7. Sustainability This project is highly sustainable for the community of Agua Caliente as it is simply an effort to

make improvements to a system originally and independently created by the citizens of the

community. By successfully building the system and continually using it over the last 30 years,

the community has demonstrated their capacity to maintain and repair their current system with

minimal outside support.

The model ram pump with the newly designed valves was constructed in Minnesota for testing

using only materials that are readily available for use in Agua Caliente. As a result, any materials

necessary for repairs to the system will be easily accessed by the community. This ensures that

the project will be highly sustainable for the community of Agua Caliente. If not for the lack of

engineering knowledge in the community, the system would be entirely sustainable without any

outside help apart from welding. For this reason EWB-USA UMN will hold a series of

workshops in Agua Caliente demonstrating the installation of new ram pump valves in order to

educate and enable community members to make adjustments and repairs to the system as

necessary. Sufficient time will be allowed for community members to ask questions to further

understand how the new design works. EWB-USA UMN will also provide the community with

detailed graphics and instruction manuals for future reference on the repair and installation of the

new valves.

This system has already proven to be financially and ecologically sustainable for the community

by their 30 year use of the system. The community has experienced significant financial benefits

as a result of the irrigation from the use of ram pumps and the installation of the new valves will

only further increase the financial sustainability of the system. The system of ram pumps in use

by the community has had minimal impact on their surrounding environment. The installation of






the new valves will not change this impact and will maintain the ecological sustainability of the

system.

7.1 Background

The community of Agua Caliente has a coop called APROMAC which grows blackberries as a

commercial crop in both wet and dry season and sells to a middle man from Miami. They are

able to do this because the water from an irrigation system they built allows them to grow higher

quality product and produce in two seasons. This allows them to skip a step in the purchasing

chain and make more money for their product rather than selling their crop in a market. The

community’s irrigation system uses a system of five irrigation dams on el Rio de Agua Caliente

to provide head to a set of ram pumps off of each dam. This allows the community to use the

river water to irrigate their crops in the dry season, providing an extra growing season in the

community. On an earlier implementation, dam 3 was determined to be in severe structural

disrepair. Dam 3 was successfully reinforced by erecting buttresses on the sides of the dam to

prevent a failure as well as adding a key and a toe onto the front of the dam. In the community

there are currently 42 ram pumps on the 5 dams, and our group has determined that these pumps

are not run at their maximum efficiency. APROMAC has asked EWB-UMN to protect and

improve their system, and therefore the step following structurally securing the dams was

working on the ram pumps. For this reason, we are testing two EWB-UMN designed valves and

a replica of the existing community valve to create a simpler, more tunable design. EWB-UMN

hopes to bring their more efficient design to the community to demonstrate the better valve so

people can use it with their ram pumps. EWB-UMN will demonstrate how to install and tune

their valves with the existing pumps so the community can build and implement the valves for

their own pumps with the knowledge from the workshops.

7.2 Operation and Maintenance

Along with the five year monitoring plan EWB-USA UMN provides, there are many things that

the community can do to ensure proper working order of the ram pumps. Before placement of

the ram pump pipes into the stream at the beginning of the dry season, it would be ideal to grind

off rust and apply a corrosion protecting paint if readily available and affordable for the

community. Another way to prevent any malfunctions of the ramp pump is to have a filtering

system at the intake of the drive pipe. This will stop any debris from entering the intake of the

pipe and flowing into the ram pump. Having a filter with the correct filter size is imperative so

that the flow of water is not impeded by debris such as twigs, leaves, plastic bottles, etc. To make

sure that the efficiency of the ram pump stays very high, maintaining the correct snifter size is

key so that the air levels in the air chamber stay consistent. If any valves of the ram pump break

and become dysfunctional, they need to be changed as fast as possible to ensure the correct

operation of the pump. The valves can be easily fabricated by the owners of the ram pumps

within the community. The pumps should also be properly tuned so that they are working at the

optimal efficiency. The ram pumps owners will be taught how to tune the ram pumps during the

workshops, so they will be able to tune them on their own.






7.3 Education

7.3.1 Workshops and Community Interaction

During this implementation trip, EWB-USA UMN will hold a series of workshops

demonstrating to the members of Agua Caliente how to properly build and tune our newly

designed valves for their ram pumps. When arriving on site, the travel team will build and run

the new impulse valve design, without community members, to ensure the valve runs as

predicted. They will then hold a workshop teaching the community how to assemble and install

the new impulse valves so the community will be able to replicate the design and process in the

future. This workshop will also explain, in detail, why it is necessary to assemble the valve

according to the provided procedure, as well as why the new design is more powerful than the

current system. An additional workshop will be held educating the community on the

importance and procedure of properly tuning the ram pumps to maximum flow rate. In addition

this workshop will demonstrate proper maintenance techniques that will ensure proper ram pump

operation. Both workshops will utilize an existing ram pump to demonstrate the improved

impulse valve. By giving special attention to the building process, the workshops will allow the

community members to have the necessary knowledge to ensure replication and sustainability of

the new valve design. By encouraging the community to follow our design and construction

techniques, we can improve the longevity and minimize down-time of the pumps and the system.

When in country, a manual, translated into Spanish, will be left with the community. This

manual will contain specific instructions on proper design, assembly and tuning procedures of

the new impulse valve. The manual will also include a list of potential complications that may

occur and how to properly resolve them. By leaving the community with the proper education

and a manual, EWB-USA UMN will provide the community the necessary tools to build and

maintain the new valves after we are gone. The time spent educating the community on proper

ram pump construction, in addition to the manual, will allow the community to reproduce the

same results and sustain the system for a long period of time.

7.3.2 Introducing New Farmers to the System

In order to introduce new farmers to the system, we will teach them how to utilize a throttle

device along with an existing pump near them. To test whether or not someone’s farm can now

utilize the system, the farmer would borrow their neighbor’s optimized ram pump. They will put

a throttle device on the delivery output which is set with the added pressure of the height

difference between the field utilizing the system and the field being potentially added. The

farmer will re-optimize the pump at the added pressure and see if it has a consistent output at the

increased pressure. If the pump outputs water with the throttle and the existing height, then the

new farmer will be able to utilize the system with the new pump design.

7.3.3 Measuring the Increased Output of Existing Pumps

For the people with existing ram pumps, we teach them the following process in order to

calculate their increased usage output:






1. Measure the flow rate of the existing valve at the height of one of their fields.

2. Exchange the existing valve for the elbow and new weight-based valve we are teaching

the community how to build.

3. Measure the flow rate of the new valve at the same height as the first valve, but with

varying stroke length and weight to figure out the maximum flow rate for the given

delivery head. This will be tested based on similar trials to the trials EWB-UMN

executed in a lab setting.

4. Run the pump at the optimum tune settings to demonstrate to the community that by

tuning the ram pump you can achieve a higher flow rate at a given delivery height with

the new valve design.

7.3.4 Introducing New Delivery Height to Community

EWB-UMN will take several steps in order to demonstrate to the community that the new valve

can reach a higher delivery height.

1. Throttle the pump in the stream bed to simulate increased delivery heights and tune the

pump for maximum output flow at each delivery head setting.

2. Bring the throttle device up to a field currently supplied by a ram pump. EWB-UMN

will measure the maximum flow rate at the given height, then throttle the output to a

higher simulated height to demonstrate that the community could pump to a higher field

than currently possible because of the ability to tune.

3. Bring the delivery hose of an existing ram pump up to a height where it can no longer

output water. Switch the existing valve to the new valve, tune the valve, and demonstrate

that the tunable valve can still output water, albeit a smaller amount of water.

4. There will be some heights where the maximum flow rate at the output is too minimal to

be useful to farmers. We will establish with the community what the minimum

sustainable daily quantity of water is, and determine where this amount is able to be

pumped to.

8. Signed Implementation Agreement While this agreement isn’t yet signed with the community, we will sign it with the leaders of the

community through our NGO Long Way Home before arrival for implementation.

Engineers Without Borders

University of Minnesota-Twin Cities Chapter

Minneapolis, Minnesota






1 March 2013

APROMAC de Agua Caliente


Regarding: The Acquisition, Improvement, Operation, and Maintenance of Ram Pumps.

This contract is an agreement on the improvement and maintenance of the ram pumps in Agua

Caliente between APROMAC and Engineers Without Borders University of Minnesota-Twin

Cities Chapter (EWB-UMN). Engineers Without Borders and APROMAC will undertake the

stated responsibilities for the acquisition, improvement, and monitoring process.

APROMAC will:

Provide 2 laborers for construction of the ram pumps.

Provide preparation for the demonstration drive pipe whether a disassembled pump from

a member of the community, a drive pipe built from spare parts, or a drive pipe purchased

and welded before EWB-USA UMN’s arrival in country.

Choose key members of the community to attend all workshops provided by EWB-USA

UMN about the design of improved ram pumps.

Be responsible for purchasing materials for and assembling their own valves to improve

their own ram pumps.

Continue having monthly contact with Engineers Without Borders about updates in the

community, maintenance of the system, and future coordination with EWB-UMN

Regularly follow the monitoring procedure provided by Engineers Without Borders to

preserve the integrity of the ram pumps and perform repairs.

Engineers Without Borders will:

Provide and cover the cost of the materials for construction of a demonstration ram pump

and build the demonstration with the help of the APROMAC workers.

Plan and provide a series of workshops demonstrating how to build, tune, and maintain

the ram pumps.

Provide educational materials for reference for the community to continue utilizing the

technology once EWB-USA UMN leaves the community.

Continue to maintain monthly contact with the community to discuss maintenance of the

EWB-UMN implementation and preparations for future coordination.

Provide the community with a maintenance plan for the sustainability and longevity of

the system.

Monitor the ram pumps for 5 years and be responsible for repairing any damage directly

resulting from design or structural flaws.






Both Engineers Without Borders and APROMAC will uphold their stated responsibilities or the

implementation of ram pumps could be at risk delayed. At the time of construction a monitoring

agreement will be signed.

________________________ ________________________

President of APROMAC Student Lead of EWB UMN

________________________ ________________________

Vice President of APROMAC Professional Lead of EWB UMN

9. Site Assessment Activities Trips prior to this have yielded quite a bit of information we have needed to perform many of our

calculations and designs. Most of this data has been quantitative like measuring the flow rate and

pressure in several of the ramp pumps on the dams. Also, we assessed all of the five downs in

Agua Caliente to see which dam was in the worst condition and needed to be repaired. These

calculations include the amount of water held behind the dam, the physical size of the dam, and

the condition seeing if there are any cracks, erosion, or seepage underneath the dam. On the

larger scale, we took GPS points around all the dams so that we could map the terrain. This was

important to calculate the water shed for the whole Agua Caliente valley to know how rain

affected the water level. Another significant part of prior assessments was to actually survey the

community, including the people in and outside of APROMAC. These questions included

whether or not they owned a pump on any of the dams, the size of their farms, and the amount of

water they used need to supply to their land.

10. Professional Mentor Assessment

10.1 Professional Mentor Name and Role

Kim Haglund, lead mentor for ram pump testing.

10.2 Professional Mentor Assessment

Three valve designs were tested in order to determine which design produces the highest amount

of delivered water. It is impractical to test the ram pumps under the same conditions that are

found in-country. Therefore, a design similar to the pump used by the community was built

along with two proposed improvements to the design. These three pumps are tested under the

same conditions so that a relative comparison can be drawn. Despite the difference in






conditions, any relative improvements that were seen during the experiment should also extend

to the usage conditions.

A designed experiment was created in order to test the effects of stroke length and valve weight

on the delivered water from each pump design. Response Surface analyses of the data generated

models for each design in order to predict the optimal values for each factor and the expected

amount of water delivered from a design that uses the optimal values. One of the valve designs

clearly showed that its predicted water delivery is higher than the other two designs, including

the design that most closely replicates the design used in-country. The group plans to implement

the corresponding design changes on five ram pumps in the community, and educate the

community on the tuning effects to the design so that they can implement the design on other

pumps and effectively maintain the new design.

10.3 Professional Mentor Affirmation

I agree with the experimental method presented in this report and I recommend the

implementation of design changes to the community’s ram pumps according to the changes that

were made to valve #1.






Appendix A – Existing System Pictures

A.1 Check Valve

Above are pictures of the check valve or one way valve of the existing ram pumps. Water comes

up through the holes and underneath the rubber flap in the picture to the right. When the

pressure underneath the valve becomes negative, the rubber quickly shuts against the holes and

no water is let through. This allows the pump to produce water up gradients.

A.2 Waste Valve

Above you can see pictures of the existing impulse valve slider rod. The design in guatemala

consists of a metal rod with a thick piece of rubber and a metal plate attached to the end. This

plate slams shut when the pressure of the water flowing past the valve and out the output

becomes greatere than the spring or weight force holding it open. Our design currently has the

same function without the rubber, creating a sharper impulse and a therefore more intense

pressure wave with less energy loss.






Above are pictures of the exsisting waste valve. You can see the four pieces of metal welded to

the end of the body of the ram, making a structure to stabalize the waste valve slider rod. There

are several loops welded to the body as well, which provide tie-offs for the rubber strips that the

community currently uses for their waste valve force.






Pictures of the functioning in-country waste valve.






A.3 Air Chamber

A view inside the air chamber of the ram. As you can see, water is forced up the small interior

pipe, and the air cushon sits in the space above the internal pipe and below the top of the air

chamber.






A.4 Exemplary Existing Pumps






Appendix B – Lab Testing

The pump being built in the shop.






Assembling the body of the ram.






The output of the ram, where the waste valve goes.






The testing facility and the drive pipe.






Nick tightening the output. For testing purposes we put a valve on the output to simulate

pressure and measured it with a pressure gauge.






Nick tightening the output valve with the pump fully assembled. In the bottom of the picture the

waste valve can be seen. The discs are the variable weights we used for the valve. The structure

on top of the flange is the guide for the slider rod.






Appendix C – Analysis of Experimental Design

VALVE #1

DATA ANALYSIS

Response Surface Regression: Delivered Water (lb.) versus Stroke (in), Valve Weight (lb.) The analysis was done using coded units.

Estimated Regression Coefficients for Delivered Water Weight (lb.)

Term Coef SE Coef T P

Constant 26.3190 1.622 16.223 0.000

Stroke (in) 4.6408 1.313 3.534 0.010

Valve Weight (lb.) 3.0264 1.273 2.377 0.049

Stroke (in)*Stroke (in) -6.1252 1.462 -4.189 0.004

Valve Weight (lb.)*Valve Weight (lb.) -4.7482 1.363 -3.484 0.010

Stroke (in)*Valve Weight (lb.) -0.4783 1.785 -0.268 0.796

S = 3.63084 PRESS = 624.748

R-Sq. = 86.21% R-Sq.(pred) = 6.63% R-Sq.(adj) = 76.36%

Analysis of Variance for Delivered Water (lb.)

Source DF Seq SS Adj SS Adj MS F P

Regression 5 576.849 576.849 115.370 8.75 0.006

Linear 2 218.942 240.688 120.344 9.13 0.011

Stroke (in) 1 144.598 164.610 164.610 12.49 0.010

Valve Weight (lb.) 1 74.344 74.467 74.467 5.65 0.049

Square 2 356.961 356.735 178.367 13.53 0.004

Stroke (in)*Stroke (in) 1 197.059 231.335 231.335 17.55 0.004

Valve Weight (lb.)*Valve Weight (lb.) 1 159.902 160.018 160.018 12.14 0.010

Interaction 1 0.946 0.946 0.946 0.07 0.796

Stroke (in)*Valve Weight (lb.) 1 0.946 0.946 0.946 0.07 0.796

Residual Error 7 92.281 92.281 13.183

Lack-of-Fit 3 90.033 90.033 30.011 53.40 0.001

Pure Error 4 2.248 2.248 0.562

Total 12 669.130

Unusual Observations for Delivered Water (lb.)

Delivered

Water

Obs StdOrder Weight (lb.) Fit SE Fit Residual St Resid

9 9 3.700 9.627 2.744 -5.927 -2.49 R

11 11 18.500 14.105 2.961 4.395 2.09 R

R denotes an observation with a large standardized residual.

- P-values show that stroke length and valve weight are both significant, but there is no interaction between

the two factors.

- R-squared value of 86.21% shows that the model represents the data well.






- Observations 9 and 11 appear to be unusual. Upon examination of the data, Observation 9 had a very low

response that was likely due to the very low stroke length. We decided to re-analyze the data without this

trial to see if the model improves.

DATA ANALYSIS (EXCLUDING TRIAL 5)

Response Surface Regression: Delivered Water (lb.) versus Stroke (in), Valve Weight (lb.) The analysis was done using coded units.

Estimated Regression Coefficients for Delivered Water (lb.)


Constant 26.5051 0.5881 45.067 0.000

Stroke (in) 2.1145 0.6007 3.520 0.013

Valve Weight (lb.) 3.0038 0.4611 6.514 0.001


Valve Weight (lb.)*Valve Weight (lb.) -5.7344 0.5139 -11.158 0.000

Stroke (in)*Valve Weight (lb.) -0.3951 0.6466 -0.611 0.564

S = 1.31483 PRESS = 99.8028




Regression 5 377.785 377.785 75.557 43.71 0.000

Linear 2 89.936 94.879 47.439 27.44 0.001

Stroke (in) 1 18.440 21.418 21.418 12.39 0.013

Valve Weight (lb.) 1 71.496 73.353 73.353 42.43 0.001

Square 2 287.204 287.207 143.604 83.07 0.000


Valve Weight (lb.)*Valve Weight (lb.) 1 215.265 215.253 215.253 124.51

0.000

Interaction 1 0.645 0.645 0.645 0.37 0.564

Stroke (in)*Valve Weight (lb.) 1 0.645 0.645 0.645 0.37 0.564

Residual Error 6 10.373 10.373 1.729

Lack-of-Fit 2 8.125 8.125 4.062 7.23 0.047

Pure Error 4 2.248 2.248 0.562

Total 11 388.158


Delivered

Water

Obs StdOrder Weight (lb.) Fit SE Fit Residual St Resid

13 13 15.000 16.711 1.048 -1.711 -2.15 R


- P-values show that stroke length and valve weight are both significant, but there is no interaction between

the two factors.

- R-squared value of 97.33% shows that the model represents the data very well.






- Observation 13 appears to be unusual, but since the data is still modeling well, we chose to include this trial

in the analysis.

210-1-2

99

90

50

10

1

Residual

Pe

rce

nt

25201510

1

0

-1

-2

Fitted Value

Re

sid

ua

l

1.51.00.50.0-0.5-1.0-1.5

3

2

1

0

Residual

Fre

qu

en

cy

13121110987654321

1

0

-1

-2

Observation Order

Re

sid

ua

l

Normal Probability Plot Versus Fits

Histogram Versus Order

Residual Plots for Delivered Water (lb)

- Normal Probability Plot: points along the line show that the data is normal

- Histogram: shows there is no evidence of skewness or outliers

- Versus Fits: random scattering of data shows no evidence of nonconstant variance

- Versus Order: random scattering of data shows no evidence of error correlation






Stroke (in)

Va

lve

We

igh

t (l

b)

2.52.01.51.00.5

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

>

–

–

–

–

< 5

5 10

10 15

15 20

20 25

25

(lb)

Weight

Water

Delivered

Contour Plot of Delivered Water (lb) vs. Valve Weight (lb), Stroke (in)

- Data points show the combinations that were tested

- Darkest green area shows optimal values

Response Optimization Parameters

Goal Lower Target Upper Weight Import

Delivered Water (lb.) Maximum 23 28 28 1 1

Global Solution

Stroke (in) = 1.77768

Valve Weight (lb.) = 2.87558

Predicted Responses

Delivered Water (lb.) = 27.1825 , desirability = 0.836494

Composite Desirability = 0.836494

- The optimal values for this valve design are a stroke length of 1.78in and a valve weight of 2.88lb. This

combination is expected to result in a response of 27.2lb of delivered water.

VALVE #2






DATA ANALYSIS

Response Surface Regression: Delivered Water (lb.) versus Valve Weight (lb.), Stroke (in) The analysis was done using coded units.



Constant 7.1071 0.5190 13.694 0.000

Valve Weight (lb.) 1.8464 0.4482 4.119 0.003

Stroke (in) 1.7168 0.4495 3.819 0.005

Valve Weight (lb.)*Valve Weight (lb.) 0.1608 0.4666 0.345 0.739


Valve Weight (lb.)*Stroke (in) 0.2534 0.6314 0.401 0.699

S = 1.27139 PRESS = 53.4689




Regression 5 51.7375 51.7375 10.3475 6.40 0.011

Linear 2 51.1738 51.0067 25.5034 15.78 0.002

Valve Weight (lb.) 1 27.5495 27.4309 27.4309 16.97 0.003

Stroke (in) 1 23.6243 23.5758 23.5758 14.59 0.005

Square 2 0.3033 0.3033 0.1516 0.09 0.911

Valve Weight (lb.)*Valve Weight (lb.) 1 0.2129 0.1920 0.1920 0.12 0.739


Interaction 1 0.2604 0.2604 0.2604 0.16 0.699

Valve Weight (lb.)*Stroke (in) 1 0.2604 0.2604 0.2604 0.16 0.699

Residual Error 8 12.9314 12.9314 1.6164

Lack-of-Fit 3 6.1087 6.1087 2.0362 1.49 0.324

Pure Error 5 6.8227 6.8227 1.3645

Total 13 64.6689

Estimated Regression Coefficients for Delivered Water (lb.) using data in uncoded

units

Term Coef

Constant 1.73379

Valve Weight (lb.) 0.520525

Stroke (in) 2.10483

Valve Weight (lb.)*Valve Weight (lb.) 0.0714681

Stroke (in)*Stroke (in) -0.196654

Valve Weight (lb.)*Stroke (in) 0.225269






210-1-2

99

90

50

10

1

Residual

Pe

rce

nt

1210864

1

0

-1

Fitted Value

Re

sid

ua

l

1.51.00.50.0-0.5-1.0-1.5

4.5

3.0

1.5

0.0

Residual

Fre

qu

en

cy

1413121110987654321

1

0

-1

Observation Order

Re

sid

ua

l













Valve Weight (lb)

Str

oke

(in

)

54321

2.0

1.5

1.0

0.5

>

–

–

–

–

< 4

4 6

6 8

8 10

10 12

12

(lb)

Water

Delivered

Contour Plot of Delivered Water (lb) vs Stroke (in), Valve Weight (lb)


VALVE #3

DATA ANALYSIS

Response Surface Regression: Delivered Water (lb.) versus Spring Tension (lb.), Stroke (in) The analysis was done using coded units.



Constant 9.2988 0.9011 10.319 0.000

Spring Tension (lb.) -0.6429 0.9869 -0.651 0.536

Stroke (in) -0.2731 0.8350 -0.327 0.753

Spring Tension (lb.)* -0.8187 0.9889 -0.828 0.435

Spring Tension (lb.)

Stroke (in)*Stroke (in) 1.7724 0.8882 1.996 0.086

Spring Tension (lb.)*Stroke (in) -0.4238 1.1835 -0.358 0.731

S = 2.23802 PRESS = 712.923









Regression 5 22.3489 22.3489 4.4698 0.89 0.534

Linear 2 0.8448 3.0885 1.5443 0.31 0.744

Spring Tension (lb.) 1 0.4499 2.1255 2.1255 0.42

0.536

Stroke (in) 1 0.3949 0.5357 0.5357 0.11 0.753

Square 2 20.8618 21.2827 10.6414 2.12 0.190

Spring Tension (lb.)*Spring Tension (lb.) 1 1.4885 3.4332 3.4332 0.69

0.435


Interaction 1 0.6422 0.6422 0.6422 0.13 0.731

Spring Tension (lb.)*Stroke (in) 1 0.6422 0.6422 0.6422 0.13

0.731

Residual Error 7 35.0611 35.0611 5.0087

Lack-of-Fit 2 20.8615 20.8615 10.4308 3.67 0.104

Pure Error 5 14.1995 14.1995 2.8399

Total 12 57.4100


Delivered

Obs StdOrder Water (lb.) Fit SE Fit Residual St Resid

2 2 8.480 9.433 2.188 -0.953 -2.03 R



units

Term Coef

Constant -2.61101

Spring Tension (lb.) 1.13772

Stroke (in) -8.56747

Spring Tension (lb.)* -0.0145546


Stroke (in)*Stroke (in) 3.15101

Spring Tension (lb.)*Stroke (in) -0.0753387

DATA ANALYSIS (EXCLUDING TRIAL #2)

Response Surface Regression: Delivered Water (lb.) versus Spring Tension (lb.), Stroke (in) The analysis was done using coded units.



Constant 8.872 0.6405 13.851 0.000

Spring Tension (lb.) 2.995 1.4157 2.116 0.079

Stroke (in) -1.788 0.7750 -2.307 0.061

Spring Tension (lb.)* 3.199 1.5308 2.090 0.082


Stroke (in)*Stroke (in) 2.347 0.6453 3.637 0.011

Spring Tension (lb.)*Stroke (in) -4.781 1.6959 -2.819 0.030






S = 1.54918 PRESS = 42.8202




Regression 5 40.5777 40.5777 8.1155 3.38 0.085

Linear 2 2.6359 14.4187 7.2093 3.00 0.125

Spring Tension (lb.) 1 0.0891 10.7431 10.7431 4.48

0.079

Stroke (in) 1 2.5468 12.7726 12.7726 5.32 0.061

Square 2 18.8705 36.3084 18.1542 7.56 0.023

Spring Tension (lb.)*Spring Tension (lb.) 1 0.2744 10.4791 10.4791 4.37

0.082


Interaction 1 19.0713 19.0713 19.0713 7.95 0.030

Spring Tension (lb.)*Stroke (in) 1 19.0713 19.0713 19.0713 7.95

0.030

Residual Error 6 14.3998 14.3998 2.4000

Lack-of-Fit 1 0.2002 0.2002 0.2002 0.07 0.801

Pure Error 5 14.1995 14.1995 2.8399

Total 11 54.9775


Delivered

Obs StdOrder Water (lb.) Fit SE Fit Residual St Resid

8 8 5.780 9.086 0.632 -3.306 -2.34 R



units

Term Coef

Constant 35.0388

Spring Tension (lb.) -2.37821

Stroke (in) 14.8829

Spring Tension (lb.)* 0.0568648


Stroke (in)*Stroke (in) 4.17253

Spring Tension (lb.)*Stroke (in) -0.849882






20-2-4

99

90

50

10

1

Residual

Pe

rce

nt

161412108

0

-2

-4

Fitted Value

Re

sid

ua

l

210-1-2-3

7.5

5.0

2.5

0.0

Residual

Fre

qu

en

cy

151413121110987654321

0

-2

-4

Observation Order

Re

sid

ua

l













Spring Tension (lb)

Str

oke

(in

)

5045403530

2.5

2.0

1.5

1.0

>

–

–

–

< 10

10 20

20 30

30 40

40

(lb)

Water

Delivered

Contour Plot of Delivered Water (lb) vs Stroke (in), Spring Tension (lb)


Appendix D – Cited Sources [1] Eshenaur, Walter C. A Theoretical Computer Based Model for Use in Design of Hydraulic

Ram Water Pumps. Thesis. University of Minnesota, 1985. N.p.: n.p., n.d. Print.

[2] Watt, S. B. A Manual on the Hydraulic Ram for Pumping Water. N.p., n.d. Web.

[3] Welch, Michael. Things That Work! The Folk Ram Pump. Home Power, n.d. Web.

document 525 pre-implementation report chapter: …

Documents