Top Banner
Water distribution network optimization by non-technical workers the application of decision support systems to analyzing peace corps data case study of el Socorro, honduras CAITLIN AUGUSTIN PHD STUDENT, UNIVERSITY OF MIAMI [email protected] LATOYA KILGOUR PHD CANDIDATE, UNIVERSITY OF MIAMI CHARLES ERGHOTT PHD CANDIDATE UNIVERSITY OF MIAMI SHIHAB ASFOUR CHAIR, DEPARTMENT OF INDUSTRIAL ENGINEERING UNIVERSITY OF MIAMI
30

Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Feb 06, 2016

Download

Documents

ursala

Water distribution network optimization by non-technical workers the application of decision support systems to analyzing peace corps data case study of el Socorro, honduras. Caitli n Augustin PhD Student, University of Miami [email protected] Latoya kilgour - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Water distribution network optimization by non-technical workers

the application of decision support systems to analyzing peace corps data

case study of el Socorro, hondurasCAITLIN AUGUSTIN

PHD STUDENT, UNIVERSITY OF MIAMI

[email protected]

LATOYA KILGOURPHD CANDIDATE, UNIVERSITY

OF MIAMI

CHARLES ERGHOTTPHD CANDIDATE

UNIVERSITY OF MIAMI

SHIHAB ASFOURCHAIR, DEPARTMENT OF

INDUSTRIAL ENGINEERINGUNIVERSITY OF MIAMI

Page 2: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Project BackgroundIn 2008 Peace Corps

volunteers enlisted UM-EWB to assist on the development of a water distribution system for the town of El Soccorro, Honduras

Student teams traveled to El Socorro five times over the next two years

Pipeline was constructed in spring 2010

Page 3: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Key PointsProblems encountered in the El Socorro

project were not unique◦The Peace Corps does not have engineers

filling engineering roles◦Decision making is haphazard◦Management of finite resources

Solutions proposed in El Socorro can be applied in all communities where Peace Corps works◦The Peace Corps uses standardized excel

forms for data collection◦A decision support system can optimize

existing water and funding resources

Page 4: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Presentation overviewProject introductionStudy tools

◦ IE Tools◦Peace Corps Data Collection

Tools◦Community Interviews

Decision Support System◦Build of Model◦Example Run

General Recommendations

Conclusion

Page 5: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Water is one of the most basic of all needs—we cannot live for more than a few days without it. And yet, most people take

water for granted. We waste water and don’t realize that clean water is a very limited resource. More than 1 billion people

around the world have no access to safe, clean drinking water—Robert alan.

Project Introduction

Page 6: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Background: water needs worldwide

Rural communities in developing countries often find it difficult to provide safe water to its residents due to costs, technical expertise, and economies of scale.

These communities rely on international aid and volunteers to assist on the projects. These volunteers often rely on 'on-the-job ' training and tools to design systems for water distribution, and lack computer software to help run initial simulations and models.

The primary motivation behind these projects is to assist the communities by helping create necessary infrastructure to meet basic water needs.

Page 7: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Project Introduction

The project seeks to provide the Peace Corps and other non-governmental organizations an assessment tool for use in region on water distribution projects

This decision support system is designed to function with Peace Corps standard topographical data collection forms

Page 8: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Project challenges Expertise- volunteers are generalists Competing Interests- community desires,

government desires Finite Resources – limited number of resources

to distribute in decision making process Long Range Planning – short terms mean high

turnover and lack of ‘big picture’ awareness Public/Private Relationships – relationship

between decision makers are often in conflict Business Proficiency – no budgeting awareness

Page 9: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

The goal is to transform data into information, and information into

insight”—Carly fiorina

Study Tools

Page 10: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Industrial Engineering Tools Risk Assessment Models: Failure Mode and

Effects Analysis, health and safety risk assessment

Statistical Quality Control: Xbar and R charts created

Cost Analysis: Created a cost analysis of the cost required to implement telemedicine in various situations.

Optimization models: Usage of population density, water source location, and budget

Decision Support Systems: Programs were designed in VBA/Excel for managers to create

Engineering Project Management: Project was organized and managed using software such as Microsoft Project, and Microsoft Visio.

Page 11: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Peace Corps Data collection

Topographical surveys◦GPS of each home◦GPS of water sources

Community mapsNumber of

people/home◦Future construction

Water quality testsWater needs survey

◦Water usage/day

Page 12: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Community interviews conducted

Water usageWater needsHealth of population

(by age group)Willingness to pay

(Contingent Valuation Method)

Current water access

Page 13: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Decision Support System

Page 14: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Model Definition

Integer programming optimization model Designed to handle large-volume data

imports (500+ data points) Provide optimized water lines based on

the cost per capita for construction The user can specify budget The user can obtain volume provided/house

and volume/capita

Page 15: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Database

Provides the data from which our solver will make a decision and allows user to access, manipulate, and query data

Imported file supplied by the Peace Corps

Page 16: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Overview: Knowledge base Our DSS model allows for storage of information

in the model through generated reports Seeks to advance the general knowledge base

by providing a preliminary design suggestion for water distribution systems designed by non governmental organizations

Page 17: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

OVERVIEW: USER

Our DSS is made for ‘managers’ rather than ‘staff specialists’

Designed to be user-friendly for a computer users of all experience levels.

Page 18: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

The solverOptimal water distribution lines based on

construction cost and environmental factorsSolver provides us with us:

◦Binary table of which pipelines to implement◦Volume of water per house for each pipeline◦Extra cost associated with water per house

Outline of how it works◦“The Main procedure is called from the Start

button, which then initializes the problem input parameters …etc.”

Page 19: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Selecting The environmental factor: Flow rate

There is a UN standard that each person needs 13.7 gallons of water a day (.051860 m³)

This constraint seeks to determine if the gallons/day requirement is satisfied, and if not, what the additional cost isVolume provided per house/day (in gallons)=

General Equation: (Cross Sectional Area)(3600)(hrs of operation)(Sqrt((head loss*2*gravity*(Pipe Diameter/Length))/(friction factor* .003785 m³))

Specific Equation: (.008105 m^2)(3600)(8 hrs)(Sqrt ((110 m*2*9.81 m/s^2)(.1016 m/LENGTH))/ (.0119*.051860 m³))

Used in conjunction with cost dataCost calculation

General Equation: 10 USD* ((13.7 gallons*6 people per house* number of houses)- (number of houses*6 people per house*gallons provided))

Page 20: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Resolve Options

User has the ability to◦Allows user to input other datasets◦Add or delete specific pipelines ◦Turn on/off environmental indicators◦Change budget amount

Page 21: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Case study El Socorro, Honduras

Page 22: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Background: el Socorro, Honduras

Current situation (2009)Approximately 1200 residents

only receive water pumped to their home every 15 days

The water isn’t treated to potable standards

They currently supplement water consumption with purchases from the nearest town—spending approximately 50% of their income on water.

Goal: design a water distribution system that will maximize existing resources to provide the maximum number of people potable water

Page 23: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

User Interface: Home

Page 24: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

User interface: Load File Data

Page 25: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Database

Page 26: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Input form

Page 27: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

The solver: cost and Environmental factors

Page 28: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

El socorro water project outcomes

Budget of $30,000214 out of 217 homes served (86%)Water tariff of $2.50 US for monthly

delivery

Page 29: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

Future DevelopmentChange additional cost (current

program demonstrates how this function will work)

Include additional constraints◦Select type of pipe◦Select pipe diameter

Provide a calculation for expected lifetime of a system

Page 30: Caitli n Augustin PhD Student, University of Miami c.augustin@umiami Latoya kilgour

conclusions