GIS for Good 1

XingGuo Chen, Kelsey Gerhart, Kim Quesnel June 6, 2014

Abstract

Energy in refugee camps is a critical issue as access to reliable electricity

dramatically improves the lives of refugees displaced from their homes and living in

temporary shelters. The goal of this project is to create a tool that the United Nations

High Commissioner for Refugees (UNHCR) Innovation Office Energy Lab Coordinator,

Sam Perkins, can use to assess the potential for solar energy systems at refugee camps

throughout Asia and Africa. While our project scope initially encompassed multiple

renewable energy resources (solar, wind and hydropower), we ultimately narrowed our

focus to only include solar power. We looked at three different solar energy

technologies- Concentrated Solar Power (CSP), Photovoltaic (PV), and Solar Cookers,

and we performed analyses for two different classifications of open refugee camps. The

camp categories were based on the accuracy of the camp location data; we did a “Broad”

analysis on 157 open refugee camps throughout Asia and Africa, and a “Detailed” spatial

analysis on 35 of the 157 camps. The 35 camps were selected based on their high

population and our ability to locate them precisely using Google Earth. We looked at

need, resource availability, and camp accessibility in determining which sites would be

GIS for Good 2

most appropriate for solar energy solutions. The specific factors that we used included

the camp population, camp demographics, refugee country of origin standard of living,

solar insolation, distance from the camps to roads and transmission lines, and biomass

scarcity near the camp. We found that PV is a more applicable technology than CSP, as it

is feasible with lower solar intensity and has more evenly distributed insolation. We

were also able to narrow in on hot spots where solar energy potential is highest, namely

Chad, Sudan, Pakistan, and Botswana. Recommendations for continued improvement of

this web mapping tool would include determining more accurate camps locations,

creating a database of NGOs and for-profit organizations working on renewable energy

projects in each country, performing a more accurate assessment of each camp's

accessibility to construction materials and equipment, determining the energy demand

in each camp, and determining most feasible solar energy technology to narrow in on

applicable ranking factors.

Background

While the most basic human

needs of water, food, and shelter are

vital for survival; the need for

energy is also a critical component

of everyday life. Refugees are not

excluded from this need, as energy

in refugee camps is used for cooking, heating, cooling, lighting, charging electronics, and

other household consumptions. Historically, energy for cooking in refugee camps has

been generated using biomass (firewood), which degrades the surrounding

GIS for Good 3

environment and creates tension with neighboring communities (1). Additionally,

gathering firewood can be dangerous for women venturing out of the camps and into

the forests to find fuel. Access to electricity for other applications such as mobile phone

charging, lighting, and fans can exponentially increase the quality of life for refugees

living in camps by allowing for increased access to education, creating the potential for

business ventures, and making refugees more comfortable in their temporary homes.

Solar powered streetlights have also been shown to drastically improve the safety of

refugee camps. The United Nations High Commissioner for Refugees (UNHCR)

Innovation office (www.unhcr.org) is looking to implement renewable energy projects

at camps worldwide to improve the quality of life for refugees while reducing the

environmental impact of the camps within the host countries.

Purpose and Scope

This project aims to aid the UNHCR Innovation office in providing relief to the

enduring energy crisis in refugee camps by providing an interactive web mapping

platform that can be used to assess which camp locations are most appropriate for solar

energy solutions. The methodology and analysis used to create the final maps are

designed to pinpoint camps where solar energy solutions are feasible and would be

most beneficial to the camp community.

Renewable energy sources exist in a variety of forms differing in size, scale, cost,

efficiency, intermittency, as well as in on-site environmental consequences that arise

over the course of construction and operation. Mapping all sources would prevent us

from completing a more in depth analysis of feasible technology. Consequently, the

project scope includes a thorough analysis of the potential for three different solar

GIS for Good 4

energy technologies- Concentrated Solar Power (CSP) systems, Photovoltaic (PV)

systems, and Solar Cookers at each of the open refugee camps worldwide.

Strengths of CSP and PV Energy Systems

Concentrated solar technology and photovoltaic panels would be the least

intrusive technologies both environmentally and politically; thus lowering the risk of

host country governmental objection. Because both forms of technology are installed

and operate above the ground surface, installation is considered reversible and thus

more favorable. More importantly, the installation of either pieces of solar technology

has historically been proven to be cost effective.

Potential Barriers of Implementing CSP and PV Energy Systems

The major issues surrounding CSP and PV include upfront costs, source security, and

the inability to meet fluctuating electricity demand.

Project Goal

The goal of this project is to create easily accessible interactive web maps for

UNHCR to use when assessing the potential for solar energy projects at refugee camps

around the globe. Additionally, the maps will be available to the public and other NGOs

with an interest in implementing solar energy projects at the designated camp locations.

Study Area

The UNHCR provided us with a dataset containing the locations of all open

refugee camps as of 2013. Unfortunately, when we mapped the camp locations using

ArcGIS, we realized that the locations were not accurate enough to conduct a high-

GIS for Good 5

resolution analysis. Consequently, we decided to manually locate and map 35 of the

most populated camps, which we would be able to analyze more thoroughly.

The "Broad Analysis" locator map seen below displays 157 open refugee camps

located within 33 countries throughout Africa and Asia. Since the accuracy of these

camp locations, provided by the UNHCR, is highly uncertain, we were limited to

performing a lower-level analysis that didn't include factors related to the camps' exact

locations. The "Detailed Analysis" locator map shows the 35 manually located camps.

These are highly populated camps for which we had accurate location data and were

therefore able to perform a more detailed analysis. Our assessment for these camps

included spatial data such as distance from camps to roads and surrounding land use,

factors that we were unable to use in the broad analysis. The locator maps show the

camp locations as icons that vary in size and are indicative of the relative population of

the camp.

GIS for Good 6

Locator Maps

Broad Analysis- Active Refugee Camp Locations UNHCR 2012 Detailed Analysis- Active Refugee Camp Locations UNHCR 2012

GIS for Good 7

Data Sources

The table below shows the data sets analyzed in ArcGIS. The data sets in bold are the ones that we used in our final analysis.

Data Source Table

GIS for Good 8

Data Processing, Analysis and Methodology

The aim of this project is to create a tool that will help the UNHCR decide where

to allocate limited resources, and we wanted to provide practical solutions and

recommendations that could be used at the end of ten weeks. While our original goal

was to assess “ALL” worldwide renewable energy potential, we soon realized that this

was not practical or feasible given our time and data availability constraints. We

initially identified and mapped multiple renewable energy technologies that could be

considered for implementation at the camps, but because of the political implications,

physical limitations, and economic costs that could potentially surface throughout

energy project development, we determined that the three most practical technologies

are Concentrated Solar Power (CSP), Rooftop Photovoltaic (PV) Solar, and Solar Cookers.

The three categories of factors that we considered in determining the potential

for renewable energy projects were need, availability of natural resources, and site

accessibility.

NEED

We assessed the "need" for energy using total camp population, demographic

data, and the refugees’ country of origin Human Development Index (HDI).

Total Population:

The total population at the open camps

varies 10 fold, from the smallest camp which

houses less than 1,300 refugees to the largest

camp, which is home to over 130,000 refugees.

With limited resources, the UNHCR Innovation

office is interested in implementing solar infrastructure that will impact a large

GIS for Good 9

population, and therefore we classified camps with higher populations as having a

higher need and therefore more potential for solar energy projects.

Camp Demographics

The data set containing the number of women and children living at each camp was

collected from the UNHCR Population Statistics site and joined to the camp locations'

attribute table. We classified camps with higher percentages of women and children as

having a higher need for solar energy solutions.

Human Development Index

The Human Development Index was used as a proxy to represent the refugees pre-

existing energy needs/demand. We assessed the need by determining the quality of life

the refugees experienced prior to being displaced, with a higher country of origin HDI

correlating to a higher need for electricity. This data was collected from the United

Nations Development Program in a .csv file and joined to the camp location attribute

table.

AVAILABILITY OF NATURAL RESOURCES

Solar:

The sun is a powerful source of energy and taking advantage of the high amount

of solar radiation found in Africa and parts of Asia through the implementation of solar

technology is a practical energy solution for refugee camps. The cost of solar

technology has significantly decreased over the years, and we do not foresee any

political barriers that could potentially surface during construction and implementation.

We looked at three distinctly different forms of solar technology- Concentrated Solar

Power, Rooftop Photovoltaic Panels, and Solar Cookers. Initially, we used worldwide

solar vector data at one degree resolution that came in the form of points, and we used

GIS for Good 10

the interpolate tool to create a raster from these points. However, we were later able to

find higher resolution data (40 km) that was in the form of polygons.

1. Concentrated Solar Power (CSP)

Concentrated Solar Power requires

Direct Normal Insolation (DNI) and is typically

implemented in large-scale projects where a

grid connection is required. CSP technologies

include parabolic trough systems, a parabolic

dish, and solar power towers. After some research, we chose to model our CSP energy

generation potential after a National Renewable Energy Laboratory case study titled,

U.S. Renewable Energy Technical Potentials: A GIS-Based Analysis (2). We mapped the

potential for concentrated solar using a trough system with a dry cooling system. When

compared to the solar power tower, a trough system is less land intensive and requires

less maintenance (i.e. cleaning, adjustments, etc.). Additionally, the dry cooling system

technology cuts the need for accessible water.

We used monthly DNI data, averaged over a 10 year period, [kWh/sq-km] to

solve for the average monthly energy generation potential at each mapped refugee

camp location, with more insolation equating to a higher potential for CSP technology.

Utilizing the capacity factor conversion table found within the aforementioned

NREL case study, we began by assigning a capacity factor for each of the monthly DNI

values found within the data set (2). Values that were below 5 kWh/sq-m/day were

assigned a value of zero, as they do not meet the minimum threshold for CSP energy

technologies.

GIS for Good 11

Class Kwh/sq-m/day Capacity Factor

1 5 - 6.25 0.315

2 6.25- 7.25 0.393

3 7.25 - 7.5 0.428

4 7.5 - 7.75 0.434

5 > 7.75 0.448

Once the Capacity Factor was assigned to each DNI value, we added a field and

used the following equation to calculate the average monthly energy potential. We

looked at the average energy potential in our analysis, but also included a graph

showing the consistency of solar radiation over the course of a year in our final web

mapping application.

Monthly Energy Potential [MWh/km2] = Power Density (MW/ sq-km) * CF * (Hours/Month)

** The power density of the trough technology modeled in our analysis (2): 32.895 MW/ sq-km

2. Photovoltaic (PV)

Photovoltaic panels require Global

Horizontal Irradiance (GHI) data as opposed to

DNI. Unlike CSP technology, PV panels are able

to take advantage and utilize diffused sunlight.

Because rooftop PV is inexpensive, scalable,

and does not require grid connection it presents itself as the most promising

option. Within our final analysis we assumed a panel efficiency of 13.5% and capacity

factor of 25%. We also used monthly data for our PV analysis, as consistent solar

GIS for Good 12

radiation throughout the year is critical for implementing solar infrastructure. In our

analysis, higher energy potential equated to a greater potential for solar energy

generation.

Monthly Energy Potential [MWh/km2] = Efficiency * CF* (MWh/km2/day)* ( # Days /Month)

3. Solar Cookers

PV was also used to model the energy potential in rooftop solar panels and solar

cookers.

Hydropower:

According to the International Energy Agency, hydropower is the most common

form of renewable technology implemented worldwide (3). Hydro power is an

attractive form of renewable energy as, unlike wind and solar, hydro power technology

has the ability to (3):

Achieve high levels of efficiency

Has greater storage capacity

Supply energy generation to meet fluctuating electricity demand

Although hydropower appears to be the most promising form of renewable energy

to date, the installation of hydro in and or near existing refugee camps is not practicably

feasible. Due to the increasing threat of water scarcity, water use and rights continue to

be an area of contention worldwide. Convincing governments in Africa and Asia to

permit the construction of mini-hydro power plants for refugee camps would be

challenging if not impossible at this point in time. As a whole, refugees located in camps

are already treated poorly and expecting the government to allow for the tampering of

local water systems with the purpose of improving upon the refugees' quality of life

GIS for Good 13

does not appear as promising as we would have hoped. While we initially mapped the

locations of rivers and streams throughout Africa and Asia, we ultimately chose not to

move forward with an analysis of hydropower technology.

Wind:

In early April, we spoke with a Stanford University professor and wind power

expert about mapping wind turbine potential in refugee camps. Through our informal

conversation, she made it clear that the only way to effectively assess the potential for

wind power worldwide would be to map wind spread versus mapping averaged speed.

Because wind speed is highly variable- differing from one kilometer to the next-

averaged data would return inaccurate data results upon analysis. With this data we

narrowed our analysis to only considering wind data collected at a height of 10 km.

Large wind turbines were not considered due to high capital costs and frequent

maintenance requirements. For instance, if a large wind turbine were to break down

(which is frequent) an expert would need to be located on-site to access the hub

(generally located 80 meters above ground). Consequently, our group only considered

small inexpensive wind turbines located on rooftops.

Upon mapping wind data (a 1km resolution point data set converted to raster by

interpolating), we discovered that the wind speeds corresponding to each of the refugee

camp locations did not present promising power return and that the poor data

resolution did not allow us to make an accurate assessment of the actual potential at the

camps. As a result, we chose to discontinue our analysis of wind potential.

GIS for Good 14

SITE ACCESSIBILITY

Roads and Transmission Lines

Since access is a critical component in

determining feasibility of constructing solar

infrastructure, we included the distance to

roads and transmission lines as a factor in our

analysis. The distance to the nearest road is an

important factor when considering how construction equipment and materials will get

to the camp. The distance to nearest transmission lines is important for mapping grid-

connected Concentrated Solar Power. Utilizing the near tool, we were able to map the

distance to nearest roads and transmission lines for each respective camp. The

transmission line data was not factored into the analysis of rooftop photovoltaic and

solar cookers, and was only assessed for camps in Africa as we could not find data for

Asia.

Biomass:

Biomass is currently the major

energy source for cooking in refugee camps.

Female refugees walk an average 5

kilometers away from the refugee camps

every day to pick up wood and other

biomass, exposing them to a high risk of being attacked and raped. Solar cookers are a

solution for this endemic problem. In our analysis of the potential for solar cookers in

refugee camps, a high biomass scarcity indicates a higher need for solar camps, thus a

higher potential ranking score.

GIS for Good 15

Elevation & Slope:

At the beginning, we wanted to use the worldwide elevation database with a

30m resolution to help us understand the elevation and slope around the camps. For

elevation, we could use tools as simple as “extract value to point” to find the elevation

for any given point in a camp. However, since the given longitude and latitude data for

the camps were not precise, and the elevation is not a critical component of solar power,

we decided not to include this data in our analysis.

We also initially thought that slope, calculated from elevation data, would be an

important factor in determining where to implement solar energy infrastructure. We

created a 10 km buffer around each camp and attempted to clip those areas from the

worldwide elevation map. However, due to slow computational processing and

inaccurate camp location data, we realized that the calculated slopes for camps could be

wrong and misleading. This analysis was, however, the motivation for us to manually

locate 35 out of 50 top populated camps whose location we could accurately determine.

Data Reclassification

Reclassification

Once we complied all of our datasets, we used ArcGIS to determine the value of each

factor (DNI, GHI, distance to nearest road, scarcity of biomass, etc) at each camp.

Then, we reclassified those values on a 0-10 scale to make them comparable. Most

values were reclassified into 5 classes by natural breaks with the values of 2, 4, 6, 8,10

except Monthly Average CSP Energy Potential, Biomass and Distance to Nearest Road.

High values of 10 indicate higher potential and suitability than low values of 2.

GIS for Good 16

Reclassification of Data

With the exception of the Distance to Nearest Road, we reclassified all of the data using

natural breaks.

Distance to Nearest Roads

Most case studies that we found did not consider the distance to roads in their

energy potential analysis. However, because some of these camps are located in isolated

areas and equipment is often not available locally, we chose to include accessibility in

our detailed analysis of the 35 camps. The ranking was designed around a study that

was completed in 2008 in which the authors set their maximum permissible distance to

the nearest road to 40 km (4). Because this study was completed for the United States

instead of Africa and Asia, we chose not to set an upper bound but rather had the

distances that fell below 40 km be ranked as "more suitable" than camp locations that

are over 40 km from a road. Consequently, rather than utilizing natural breaks we

chose to manually enter in values for reclassification.

Reclassification of Biomass

GIS for Good 17

The land use database contains different land use type such as “water bodies”,

“forests” and “shrubland”, which are recorded as unique number values ("old value") in

the attribute table. We then classified those values on a scale of 0-10 based on the

scarcity of biomass corresponding to each land use type. The higher biomass scarcity

the land use type indicates, the higher reclassified value. For example, “water bodies”

are reclassified with a value 10 since no biomass is available for consumptive use while

“closed forests” are given value 1 because of wide access to biomass. The detailed

reclassification for land use can be found in the right chart. Once we reclassified the

values, we determined the percentage of each biomass type within a 5km buffer of our

camp, and calculated the overall biomass scarcity value for each camp.

Reclassification of Biomass

Rankings and Weight

GIS for Good 18

Once each attribute was reclassified, we combined and weighted the indexed

parameters to determine one single ranked value describing the relative energy

potential at each camp for the following 5 categories:

CSP energy potential at 157 camps using a broad analysis

PV energy potential at 157 camps using a broad analysis

CSP Potential at the 35 most populated camps using a detailed analysis

PV Potential at the 35 most populated camps using a detailed analysis

Potential for Solar Cookers at the 35 most populated camps using a detailed

analysis.

We took each parameter (for example, distance to nearest road or camp population)

that had been reclassified from 0-10, with 10 indicating that that attribute contributed

to a higher potential and a zero indicating that it didn't contribute at all, and combined

them using a weighted model to rank each refugee camp according to its energy

potential accounts for the following factors*:

Population: size, density, demographics

Environment: solar radiation, biomass, distance to roads, distance to

transmission lines

Social/Economic: The Human Development Index of the refugees country of

origin

Political**: countries past and present energy initiatives/policies

* The factors considered in the "broad" versus "detailed" analysis can be found in the Data Processing,

Analysis, and Methodology section

** Not used in the final ranking but made readily available for the users of the interface

GIS for Good 19

Since solar isolation is the most important aspect of a successful solar energy

system, we assigned the ranked energy potential values as 60% of the final weighted

value. For CSP, every camp that had been ranked as having a solar energy potential of 0

was also given a final score of 0. This means that the camp is not appropriate for a CSP

solar energy system.

For the Broad Analysis, we didn't know the exact locations of the camps and were

therefore limited to using solar data, camp population and demographic data,

and information on the refugees’ home country quality of life in our rankings. For the

Detailed Analysis of the 35 most populated camps, we were able to take into account

solar data, camp population and demographics, biomass near the camp, and the

distance to the nearest transmission line (Africa only) and road. For example, as seen in

the table below, the CSP Energy Potential using a broad analysis was calculated using

the following formula:

[CSP Energy Potential] = [Average Monthly CSP Energy Potential Ranking]*0.60 + [Camp

Population Ranking]*0.05+ [Percentage of Women Ranking]*0.05+[Distance to Nearest Road

Ranking]*0.05 + [Country of Origin HDI Ranking]*0.25

GIS for Good 20

Combining Reclassified Parameters to Rank Energy Potential at the Camps

GIS for Good 21

Interactive Maps

We have created five different interactive maps displayed within two mapping

web applications.

Broad Analysis: 157 Refugee Camps

This map shows the relative potential for Concentrated Solar Power (CSP) and

Photovoltaic (PV) solar energy solutions at 157 UNHCR refugee camps that are open

and active as of June 2013. The rankings are relative, with large icons indicating camps

that have more potential than those with smaller icons. Red icons on the left side of the

map indicate that CSP solutions should not be considered. The background layer

represents the average monthly solar radiation (Direct Normal Irradiance for CSP and

Global Horizontal Irradiance for PV) throughout the year in Africa and Asia, with lighter

colors representing lower intensity and darker colors showing higher intensity.

Detailed Analysis: 35 Refugee Camps

This map shows the relative potential for Concentrated Solar Power (CSP)

solutions, Photovoltaic (PV) solar energy solutions and Solar Energy Cookers at 35

heavily populated UNHCR refugee camps that are open and active as of June 2013. The

GIS for Good 22

rankings are relative, with large icons indicating camps that have more potential than

those with smaller icons. Red icons on the leftmost map side indicate that CSP solutions

should not be considered. The background layer represents the average monthly solar

radiation (Direct Normal Irradiance for CSP and Global Horizontal Irradiance for PV and

Solar Cooker) throughout the year in Africa and Asia, with lighter colors representing

lower intensity and darker colors showing higher intensity.

Conclusions

The web applications allow us to pinpoint areas and camps with a high potential

for solar energy technology and to do a side-by-side comparison of the final results. The

light pink shading seen on the maps below indicates areas that contain multiple camps

with high suitabilities. The charts below each map show the top 5 camps with highest

potential for each type of solar energy technology.

GIS for Good 23

Broad Analysis: 157 Refugee Camps, Hot Spots and Camps with the Most Potential

GIS for Good 24

Detailed Analysis: 35 Refugee Camps, Hot Spots and Camps with the Most Potential

GIS for Good 25

From our analysis, we have concluded that PV tends to be a more applicable solution

than CSP due to the fact that: it requires a lower baseline of insolation, camps can be

more isolated/less accessible for PV solutions, and PV is a more feasible and mature

technology. Looking at the maps we can also point to the hot spots that have a high

solar power development potential, namely Chad, Sudan, Pakistan and Botswana.

Next Steps

After ten weeks of working on this project, we are excited to hand over our project

to the UNHCR and anxious to see how they will use our mapping application to make

informed decisions regarding the implementation of solar energy projects. Although we

are very pleased with our final product, we envision the potential for this project in the

future. Looking to the future, we suggest incorporating the following factors into the

existing analysis:

(1) More Accurate Camp Locations

(2) Create a Database of NGOs and For-Profit Organizations Working on Renewable

Energy Projects in Each Country

(3) More Accurate Assessment of Accessibility to Construction Materials and

Equipment

(4) Determine Energy Demand in Each Camp

(5) Determine most feasible solar energy technology to narrow in on applicable

ranking factors

GIS for Good 26

Sources

(1) Lyytinen, Eveliina. Household energy in refugee and IDP camps: Challenges and solutions for UNHCR.

UNHCR, Policy Development and Evaluation Service, 2009.

(2) Lopez, Robert, et al. U.S. Renewable Energy Technical Potentials: A GIS-Based Analysis. NREL. National

Renewable Energy Laboratory. 2012. http://www.nrel.gov/docs/fy12osti/51946.pdf

(3) Renewable Energy Essential. IEA. International Energy Agency, 2010.

http://www.iea.org/publications/freepublications/publication/Hydropower_Essentials.pdf

(4) http://www.ii.umich.edu/UMICH/asc/Home?About%20Us/Documents/Ramde_Article.pdf