Professional Documents
Culture Documents
Insufficient Water Supply in Tegucigalpa
Insufficient Water Supply in Tegucigalpa
Insufficient Water Supply in Tegucigalpa
December 2011
ii
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
A CKNOWLEDGEMENTS
This research was financed by the Department of Research Cooperation of the
Swedish International Development Agency (SIDA/SAREC) within the
collaboration program between the Universidad Autónoma de Honduras
(UNAH).
I would like to thank my supervisors Jan Erik Gustafsson, Patricia Phumpiu
and Gladis Rojas for providing me with the support to carry out my research
and for helping me get started in the scientific world.
I wish to extend my profound gratitude to Berit Balfors and Stella Lowder
whose support, advice and guidance were essential for the finalization of this
research.
I wish to express my gratitude for the support provided by UNAH, especially
Lelany Pineda; Mónico Oyuela; Saul Jimenez and also to Patricia Hernandez
for providing me with working facilities during my visits to Honduras.
This research would not have been possible without the support of the
submanager of SANAA, Ricardo Velasquez and everyone who contributed
during the interviews. Special thanks to Alejandro Somoza, Rodolfo Ochoa
and Luis Romero for providing me with generous amount of information.
I am deeply grateful for the support received from the University of Glasgow,
specially Trevor Hoey and John Briggs for opening me the doors of the School
of Geographical and Earth Sciences and Maggie Cusack and Joanne Sharp for
continuing the support.
From KTH I would like to thank Aira Saarelainen, Katrin Grünfeld and Jerzy
Buczak. Your help goes far beyond the call of duty. You were always there to
provide me guidance, advice and a friendly smile whenever I needed one.
Thanks to the rest of the staff and students for the very useful discussions.
I also want to thank for the contributions of many of my friends: Alessa
Geiger, Delia Gheorghiu and Flavia Forte who helped me to elaborate the
maps and diagrams; Nelly Rodriguez who helped me to acquire programming
skills; Ida Westerberg who helped me with the hydrologic analysis and maps;
Elias García who provided very useful pictures and Garfield Tait who is
permanently helping me to improve my English.
My deepest gratitude also goes to all of my friends who have helped me to feel
like home, wherever I am. Special thanks to Zhihong Zhao who has always
been willing to help me unconditionally in all situations.
I want to thank Marita Wingren and Veronica Melander for believing in me
and giving me the opportunity to continue with my research.
Special thanks for the support of my husband Uwe Balthasar, my friend Anneli
Häyrén Weinestål, and my family. Your faith and confidence in me is what has
kept me standing all this time. Thanks for always believing in me.
iii
Zairis Coello Midence Balthasar TRITA LWR LIC 2058
iv
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
v
Zairis Coello Midence Balthasar TRITA LWR LIC 2058
vi
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
TABLE OF C ONTENT
Acknowledgements iii
Abbreviations and symbols v
Table of Content vii
List of Papers vii
Abstract 1
Introduction 1
Description of the Study Area 3
Developed Water Supply Projects 6
Temporary Solutions 8
Proposed Water Supply Projects 10
Decision Making and the Water Sector Reform 10
Methodology 13
Precipitation Analysis (Paper I) 13
Analysis of the Reasons Leading to an Insufficient Water Supply (Paper II) 15
Results and Discussion 16
Quantifying the Available Water Resource (Paper I) 17
Investment for a Sufficient Water Supply (Paper II) 19
Conclusions and Future Research 22
References 24
Other references 27
L IST OF PAPERS
I. Westerberg, I., Walter, A., Guerrero, J.-L., Coello, Z., Halldin, S., Xu, C.-Y., Chen, D. and
Lundin, L.-C. (2010). Precipitation data in a mountainous catchment in Honduras: quality
assessment and spatiotemporal characteristic. Theoretical and Applied Climatology 101(3):
381-396.
II. Coello-Balthasar, Z., and Balfors, B. (2011) Problems faced by a national water utility
in an urban area, case study: Tegucigalpa Honduras. Submitted to the International
Journal of Sustainable Development and Planning, October 2011.
vii
Zairis Coello Midence Balthasar TRITA LWR LIC 2058
viii
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
A BSTRACT
Tegucigalpa, the capital of Honduras, has experienced an unsatisfied water
demand during the last three decades. The state owned water utility in charge
of the water supply of the country, SANAA, has faced this deficit by providing
an intermittent water supply. The intermittent water supply has increased the
gap between the rich and the poor, who cannot afford water storage facilities.
Theories explain water scarcity either by low precipitation or by lack of
investment in water structures. This thesis investigates the applicability of both
explanations by quantifying the annual precipitation in the sub catchments with
water supply potential for Tegucigalpa, and identifying the problems which
caused the lack of investment into the water infrastructure. The analysis
concluded that even if the annual precipitation is abundant, it is not evenly
distributed in time and in space. Furthermore, it is argued that the financial
limitations which hindered the lack of investment in water structures originated
in the low tariffs imposed, and to the practices of the patronage system.
1
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
The intermittent water supply poses several severe as just providing water during 6 to 8
problems which have been addressed by hours on every second or third day
researchers. Vairavamoorthy (2007) listed (SANAA, 2010).
problems such as: (a) low pressure in the
A portion of the population of Tegucigalpa
network, caused by severe pressure losses
managed to cope with the intermittent
from big flows in undersized networks since
service by building or installing storage tanks
water is being consumed for a short
that are filled when water is provided, thus
duration; (b) inequitable distribution of
making the household more independent
water, since consumers in high pressure
from interruptions of supply (Fig. 2). In
areas draw more water than those in low
1992, it was estimated that households in
pressure areas; (c) contamination of water
Tegucigalpa had invested between 3 US$
occurs in networks where the pollutants
and 9 million for tanks and pumping
enter through leaks in the supply pipes that
systems to compensate for the unreliable
are empty during the prolonged periods of
supply (Lauria, 1992). However, 15.1% of
interruption of supply; (d) coping costs of
the inhabitants of Tegucigalpa live below the
the consumers, including storage facilities,
line of poverty (World Bank, 2006) and are
alternative water supplies, pumping,
not able to afford water storage facilities1. It
treatment facilities.
has been estimated that the poor would have
The intermittent service has increased social to pay 8.37 US$ per cubic meter of water to
inequality because the economic capability private water vendors and would only be
of a household to invest in the construction able to consume approximately 24 litres per
or installation of a storage facility will person per day 2(Nauges and Strand, 2007),
represent the difference between having which is less than half of the 50 litres per
water or not. Additionally, pathogen person per day recommended as the basic
contamination occurs in the stored water at
household level due to improper storage and
handling (Jensen et al, 2002). Therefore,
considering the overall poor quality of the
water supply, it is no surprise that even if the
cases of water borne diseases, such as
diarrhoea, have decreased (Deal et al, 2010),
diarrhoea still ranks as the first cause of
morbidity in Honduras, with 219,246
reported cases during 2010 (OPS, 2010). It
was also the cause of 10% of the deaths of
children below 5 years in 2008 in the whole
country (OMS, 2010).
The National Autonomous Service of Fig. 2: Installment of high storage tanks
Aqueducts and the Sewage Systems (Servicio to cope up with the intermittent water
Autónomo de Acueductos y Alcantarillados, supply has become a popular practice in
SANAA), has been the public authority Tegucigalpa (photo by Elias García).
responsible for water supply and waste water
in Honduras since 1961. Therefore, SANAA
is responsible for the water supply of 1
According to the World Bank, extreme poverty
Tegucigalpa, the capital of Honduras, with lines considers extreme poverty line as the
more than one million inhabitants (INE, monthly cost of food that would provide 2,200
2007). It was during the 1980‘s that SANAA calories per day based on observed eating
introduced an intermittent water supply in patterns. The poverty line was estimated by
adding to the costs of the extreme poverty line
Tegucigalpa in order to cope with the
a complement of non alimentary consumption.
permanent shortage (Caballero, 1992). 2 3
Based on the original estimate of 0.7 m per
Currently the intermittent service can be as month
2
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
water requirement standard for human regional level, by analyzing the rainfall
needs (Gleick, 1996). Studies have revealed distribution in time and space in the sub
that the inhabitants of the poor catchments with potential to supply water
neighbourhoods in Tegucigalpa consider for Tegucigalpa (Paper I); (ii) to detect the
that water provision is their most serious reasons that lie behind the lack of
problem (Ortiz, 2002). investment in water supply projects
(Paper II).
When Seckler et al (1999) analysed the water
supply and demand for 118 countries, Description of the Study Area
Honduras was classified in the group which The subject of analysis of this thesis is
was considered to have sufficient potential Tegucigalpa, the capital of Honduras,
water resources to meet needs by 2025, but located in the centre of the country (Fig. 1).
needed to embark on massive water However, to analyse water availability, a
development projects to achieve this different framework of analysis is required,
objective. However, that research was that is, the catchment in which Tegucigalpa
limited by the poor quality of the lies. A catchment is defined as the area of
international data on water and also its land within which all water flows to a single
availability only at a country level, not at the river system (Heathcote, 1998). The
level of internal regions. catchment containing Tegucigalpa is the
The general objective of this thesis is 7,500 km2 Choluteca River catchment
therefore, to achieve a better understanding (Fig. 3).
of the causes which lead to drinking water Figure 3 illustrates the complexity of the
scarcity in Tegucigalpa and answer the area referred to as Tegucigalpa. Tegucigalpa
research question of why Tegucigalpa does is in reality comprised of two neighbouring
not possess a continuous water supply. This cities, Tegucigalpa and Comayagüela,
objective will be achieved by (i) quantifying separated only by the Choluteca River.
the overall availability of water resources at a
3
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
Tegucigalpa was declared the capital of Cabeceras Subcatchment, also known as the
Honduras in 1849. In 1898, Comayagüela Upper Catchment. The Upper Catchment
was declared part of the capital mainly to has a total extension of 816 km2 and is
facilitate the location of diplomatic bodies located between longitudes 8728‘ and
and governmental agencies, which needed to 8705‘ W and latitudes 1410´ y 1355‘ N.
be located inside the area considered as the This Subcatchment is important for the
capital (DGEC, 1980). The National water supply of Tegucigalpa because all of
Congress declared that the union of both the existing sources, as well as the potential
cities would be considered as two new sources of drinking water which would
municipalities, forming what is known as the not require pumping, lie within this area.
Municipality of the Central District,
The Italian Consultant Company, C. Lotti &
administered by the Metropolitan Council
Associati, assessed four subcatchment areas
(Consejo Metropolitano del Distrito
in the Upper Catchment where potential
Central). Because of this administrative
new water sources could be located. These
term, Tegucigalpa and Comayagüela are
areas are: the 193 km2 Guacerique River
sometimes referred to as the Metropolitan
Subcatchment, the 141 km2 Grande River
area. For the purpose of this research, no
Subcatchment (also known as Concepción),
distinction will be made between
the 64 km2 Tatumbla River Subcatchment
Tegucigalpa and Comayagüela. The so called
and the 48 km2 Sabacuante River
Metropolitan area would be referred to as
Subcatchment (Fig. 4) (SANAA, 2005).
Tegucigalpa.
Their description of these subcatchment
Tegucigalpa is located in the upper part of areas includes the following points:
the Choluteca River Catchment (Fig. 3). The
Choluteca River catchment has an area of Highly mountainous topography with
7,500 km2. Its largest subcatchment is the elevations ranging from 900 to 2,400 meters
above sea level.
Fig 4. Lotti analysed four subcatchments, enclosed within the black lines of the
diagram, in the Upper Catchment of the Choluteca River. These subcatchments are
Guacerique (NW), Concepción (SW), Tatumbla(NE) and Sabacuante (SE) (SANAA,
2005).
4
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
5
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
(Ortiz, 2002). This growth is extremely high settlements lacking access roads, green areas,
compared to the average of 2.7% the rest of schools and health clinics (Pearce-Oroz,
Latin American capital cities had during the 2005).
period between 1975-2000 (Cohen, 2004). It is important to note that water provision
A complication in the accelerated growth of is a problem for all the neighbourhoods of
Tegucigalpa is that the city is located at a site Tegucigalpa, legal and illegal. SANAA
not well suited for urban expansion. It is provides water trucks to settlements without
completely surrounded by steep mountains access to the pipe network according to a
with very few valleys. The urban expansion calendar. The rest of the city within the
has followed the trend of irregular land piped network also receives water, according
tenure informal settlements, growing at the to a publicised calendar.
edges of the urban core in the areas with
Developed Water Supply Projects
excessive slope or near the banks of the
Choluteca River. These areas are accessible Besides the poverty of the country, SANAA
to the urban poor due to their low value and has yet another problem when trying to
undesirable topography (Pearce-Oroz, comply with their mission to provide water
2005). Dwellers in these areas face life- for the population of Honduras. This
threatening situations, mainly during the problem is rainfall seasonality composed of
rainy season, when the slopes become prone two well defined seasons, rainy and the dry.
to landslides, and settlements near river The way to guarantee water availability
banks are prone to flooding. throughout the year in these conditions is to
store water during the rainy season in
Informal settlements are trapped in a vicious
reservoirs.
circle where they need legal land titles to
qualify for basic services (i.e. water, The same year SANAA was created, it
sanitation, electricity, waste collection, etc), realized that Tegucigalpa needed a reservoir
but where service installation is a (SANAA, 1983) to guarantee water supply
requirement to obtain a title (Ortiz, 2002). In for the 164,941 inhabitants (DGEC, 1980).
1998, 225 of the 340 neighbourhoods of In May of the same year the construction of
Tegucigalpa were considered illegal a new water reservoir in the Guacerique
River was proposed as the best solution and
6
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
in 1972 SANAA decided to build the It was not until 1991, when the deficit of
reservoir. However, the project faced water had reached 37% during the dry
financial difficulties and it never came to season (PCI, 2001) and the population had
fruition, thus leaving Tegucigalpa with an increased to more than 576,661 (DGEC,
unsatisfied water demand (SANAA, 1983). 1989), that an additional source of water was
In 1973, 12 years after the Guacerique incorporated to the supply of Tegucigalpa.
reservoir had been proposed for first time, This new reservoir, Concepcion, was built
the situation was becoming so critical that by the Italian construction company Astaldi,
the ―Emergency Plan Los Laureles‖ was with Italian finance. Although Concepcion
initiated for the construction of a water was supposed to cover the water demand
reservoir for Tegucigalpa. The resulting until 2004 (SANAA, 1990), by 1998 SANAA
reservoir, Los Laureles, was finished in 1976 was forced to re-introduce an intermittent
with financial support from the Central water supply (PCI, 2001).
American Bank for Economic Integration Since the construction of Concepción, no
(BCIE) (OIRSA, 2007). The reservoir was considerable sources of water have been
built downstream of what would have been added to Tegucigalpa‘s water supply while
Guacerique and with a smaller storage the population has almost doubled and was
capacity than the original plans of 1961. This close to one million in 2007 (INE 2007). Of
diminished capacity, together with an the nine projects and the maintenance to the
increased population, meant that soon after distribution network planned in 1980 (see
its construction Los Laureles could not Fig. 5), only four have been developed. The
satisfy the water demand of the 305,387 distribution network received maintenance
inhabitants (DGEC, 1980). As a result of the only after it was partially destroyed by
continuing water shortage SANAA Hurricane Mitch in 1998 (JICA, 2006).
introduced an intermittent water supply Groundwater has been added to the water
during the 1980‘s (Caballero, 1992). supply, but only in small quantities
7
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
(Berrios, 2009). Concepcion reservoir now that whenever a new project has been
receives additional water from the Ojojona incorporated to the water supply of
River and floodgates were built above the Tegucigalpa, it has been as a reaction to an
spillway increasing the storage capacity of existing (Coello-Balthasar et al, 2011).
the reservoir by 3.0 Mm3 (JICA, 2006). In
Temporary Solutions
Los Laureles, a 3.5 m high inflatable curtain
was incorporated in 2002 to increase water In the meantime SANAA tries to find
storage by 3.5 Mm3. More water has been finance for the construction of a new
directed to Picacho and in 2010 the capacity reservoir; they have also to find ways to
of water treatment plants was increased by manage the actual deficit. The shortage of
200 l/s. Nevertheless, the storage capacity of the drinking water supply is felt more
Los Laureles is decreasing by 2% annually as intensively during the dry season. It is then
the reservoir is silting up (Lee, 1996). The when then the reservoirs fall to their
inflatable curtain, which is about to reach minimum level and the availability of water
the end of its life, has not operated in 2010 depends on whether the reservoirs stored
(Morales and Garcia, 2010). Table 1 enough water during the last rainy season.
summarizes the existing projects with their The first measure SANAA takes to prepare
improvements. Figure 6 gives an itself for the coming dry season is to create a
approximation of the location of each strategic plan for the extraction of water
source. A new project, an update of the from the reservoirs. This strategic planning
previously proposed Guacerique, now called is done by creating models of different
Guacerique II, is a priority for the SANAA scenarios that would assure the availability
management as a solution for the continuing of the stored water until the beginning of
water crisis in Tegucigalpa. the new rainy season (Table 2). The variables
This historical context helps identify the involved are extraction of water, level of
causes that led to the present crisis. History water in the reservoirs, expected demand
reveals that this is not the first time and expected beginning and intensity of the
Tegucigalpa faces water shortage. The rainy season. The models are recalibrated
pattern to introduce a new project has been continuously once the dry season starts.
Table 2: Example of one scenario for the planning of the dry season of 2006.
Translated from Zepeda (2006); Abbreviations are: Elev.=elevation; PIC = El
Picacho; MIR = Miraflores; Q=production; Vol.=stored volume;
DIVISION METROPOLITANA
PROGRAMMED PRODUCTIONS FOR 2006
PROGRAMMED PRODUCTION (m3/s)
TOTAL DEMAND DEFICIT
CONCEPCION LAURELES PIC. MIR.
Q Elev. Vol. Q Elev. Vol Q
Month Q (m3/s) m3/s m3/s m3/s
(m3/s) (m) (m3x106) (m3/s) (m) (m3x106) (m3/s)
NOV/05 1.20 1155.30 33.54 0.75 1031.70 9.15 0.90 0.065 2.92 3.25 0.34
DEC/05 1.20 1154.10 31.16 0.67 1029.40 7.53 0.90 0.065 2.84 3.25 0.42
JAN 1.10 1151.97 28.22 0.61 1025.50 5.91 0.85 0.050 2.61 3.25 0.64
FEB 1.10 1150.29 25.56 0.50 1021.98 4.54 0.75 0.040 2.39 3.25 0.86
MAR 1.20 1147.70 22.34 0.50 1018.08 3.24 0.50 0.035 2.24 3.25 1.02
APR 1.20 1144.91 19.23 0.50 1012.13 1.95 0.30 0.020 2.02 3.25 1.23
MAY 1.20 1141.52 16.02 0.50 1004.80 0.65 0.50 0.040 2.24 3.25 1.01
JUN 1.10 1138.49 13.17 0.80 1033.00 10.50 0.75 0.050 2.70 3.25 0.55
JUL 1.10 1135.13 10.22 0.80 1033.00 10.50 0.75 0.050 2.70 3.25 0.55
AUG 1.10 1130.90 7.28 0.80 1033.00 10.50 0.85 0.060 2.81 3.25 0.44
SEP 1.10 RAIN 0.80 1033.00 10.50 0.90 0.065 2.87 3.25 0.39
OCT 1.10 0.80 1033.00 10.50 0.90 0.065 2.87 3.25 0.39
8
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
9
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
10
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
Quiebramontes
Table 5: Projects proposed by
Reservoir
212,000,000 0.98 2,510
SOGREAH to solve the deficit of water
Sabacuante
in Tegucigalpa. Construction costs
Reservoir
93,000,000 0.24 4,574
include the water treatment plant but
exclude the conduction and distribution
Tatumbla
Reservoir
115,000,000 0.21 6,431 expenses associated with the new project
(SOGREAH 2004).
Reduction of Supply
AMUC
leakages by Project Cost (US$) Capacity
(US$/m³)
replacement of 160,000,000 0.32 5,749 (m3/s)
pipelines of 8” Guacerique II 132,100,000 0.58 0.87
of diameter
Del Hombre
95,400,000 0.79 0.57
Reduction of River (6)
leakages by
Del Hombre
replacement of 167,300,000 1.5 0.65
22,800,000 0.03 10,238 River (7)
pipelines of
3/4” of
diameter Sabacuante 82,800,000 0.2 1.25
11
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
Fig. 9: Diagram illustrating the location of the alternative new sources proposed by
SOGREAH and PCI.
who decides what is done? When SANAA municipalities became responsible for
was created in 1961, it was meant to have constructing, administering and maintaining
the responsibility for planning, designing, the networks of piped water and sewage,
constructing, operating, administering and functions that were the responsibility of
providing maintenance to the water supply SANAA. Therefore, the Municipal Law
and sewage systems in Honduras. The role resulted in duplicity of functions; this was
of administrator and planner of SANAA has clarified in 2003 when the Water Supply and
been criticized because of a biased Sanitation (WSS) Sector Framework Law
distribution of financial resources favouring (Ley Marco del Sector Agua Potable y
the urban areas (Walker et al, 2000). During Saneamiento) was approved.
the years to come, the role SANAA has The WSS Sector Framework Law produced
played in the water sector will radically the following changes in the water sector:
change due to changes in the legislation at a The creation of the National Water and
National level which has taken place during Sanitation Council (Consejo Nacional de
the last decades. Agua Potable y Saneamiento, CONASA).
A reform of the water sector took place in CONASA is responsible for establishing and
1990 when the National Congress of approving policies, developing strategies,
Honduras passed the Municipal Law bill. defining objectives and goals, and
The objective of this Law was the investment programs, as well as coordinating
devolution of power from the national to the activities for improving the water and
municipal levels to improve the overall living sanitation sector at a national level.
standard of different municipalities since
The creation of the Regulatory Entity for
they faced different problems which needed
the Potable Water and Sanitation Sector
local solutions. As part of this law,
(Ente Regulador de los Servicios de Agua
12
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
13
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
Ministry of Natural Resources and missing daily values: the commonly used
Environment (Secretaría de Recursos inverse-distance-weighting method (IDW)
Naturales y Ambiente, SERNA), National using squared distance and the coefficient-
Autonomous Service of Aqueducts and of-correlation- weighting method (CCWM).
Sewage System (Servicio Autónomo The latter is based on the same principle as
Nacional de Acueductos y Alcantarillados, IDW but correlation coefficients are used
SANAA), National Weather Service instead of distance to calculate the weights
(Servicio Meteorológico Nacional, SMN), of the n surrounding stations (Equation 1),
National Autonomous University of
Rmi
n
Honduras (Universidad Nacional Autónoma m i 1 i
(1)
de Honduras, UNAH) and the National
n
R
i 1 mi
Power Company (Empresa Nacional de
Energía Electrica, ENEE). The USGS (U.S. where θm is the missing value to be patched
Geological Survey) operates some automatic and Rmi is the coefficient of correlation
weather stations in co-operation with between the i:th station and station mi
SERNA. Daily data were collected from the missing a value. All available simultaneous
above-mentioned institutions. Monthly data data for the stations were used for the
were retrieved from the Global Historical calculation of the correlation coefficient,
Climatology Network (GHCN) version 2 with a minimum of 730 concurrent days set
beta dataset (Vose et al, 1992) from NOAA as a requirement. In the case of the IDW
(the U.S. National Oceanic and Atmospheric method the inverse squared distance
Administration) and from SERNA. Only between station m and station i replaces Rmi
data from the Global Historical Climatology in equation 1. The mean absolute error
Network (Vose et al, 1992) have undergone (MAE), the mean relative error (MRE), the
serious quality control. A great part of the root-mean-square error (RMSE) and the
data was reformatted into time series before coefficient of determination (R2) were used
it was put into the database. Coordinates of to evaluate the performance of the two
most precipitation stations were checked in methods at all stations with data for the two
the field while, at the same time, we periods 1985–1995 and 1996–2005. Two
investigated the measurement equipment periods were used to account for possible
and procedures. differences in performance because of
different spatial distributions of data. The
Several specific error types were best method was identified by calculating
encountered and quality-control routines and comparing the mean absolute error
were developed to tackle these errors during (MAE), the mean relative error (MRE), the
the second step. A combination of root-mean-square error (RMSE) and the
automated tests and visual data inspection coefficient of determination (R2). The
was used in this study since the amount of selected method was later used to patch the
data was not very large. Four types of quality dataset. Gaps were only patched if they were
control were performed: shorter than a month and if there was at
Outliers check for all daily values greater least one day with data during any given
than 100 mm. month.
Check for too-frequently occurring data. Spatial interpolation was performed on a
Check for sequences with too-low data. monthly time scale as more monthly data
were available and since monthly
Check for dry months during the rainy correlations were expected to be much
season. higher than daily correlations. The mean
Interpolation was then performed in the annual and mean monthly spatial
third step, where two methods for filling precipitation distributions were also
short gaps in the data were first evaluated obtained through interpolation to
and the best method was used. Two characterize the spatial properties of the
methods were evaluated for the patching of precipitation regime in the basin.
14
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
Finally, three spatial interpolation methods nugget, i.e., if there is no increase in semi-
were used, all based on weighted linear variance with distance all stations get the
combinations of the gauged data; a simple same weight. Both universal and ordinary
method – inverse-distance weighting (IDW) kriging were performed as block kriging with
– and two computationally demanding a block-size of 900 metres and a
geostatistical methods - ordinary kriging discretisation size of 100 points per block.
(OK) and universal kriging (UK) with
Analysis of the Reasons Leading to
coordinate base functions. As in the study
by Goovaerts (2000), an omni-directional an Insufficient Water Supply
semi-variogram was used because of the (Paper II)
limited number of available stations, and the The research question was approached using
variogram model was exponential with a the case study methodology. Case studies
nugget effect. In the case of universal methods are preferable when ―how‖ and
kriging, a trend is modelled as a function of ―why‖ questions are being asked about
the coordinates and subtracted where after contemporary events over which the
the residual semi-variogram is calculated. investigator has little or no control (Yin,
The Gstat geostatistical software (Pebesma 2009).The case study research method
and Wesseling, 1998) and the Gstat package requires data collection techniques using
in R (Pebesma, 2004) were used for the multiple sources of evidence to provide
interpolation. The automatic fitting of cross-data validity checks and test for
curves to the sample semi-variograms were consistency, thus performing triangulation
performed in Matlab to improve estimation (Patton, 2002; Yin, 2009). The sources of
of the nugget parameter. The ordinary information for this research were obtained
kriging reduces to a simple averaging of data by the use of qualitative data, that is, open-
if the semi-variogram is modelled as a pure ended interviews and documentation.
Fig. 10: Sequence of the interviews performed in SANAA following the snowball
sampling technique.
15
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
Qualitative methods facilitate study of issues according to Patton (2002), the sampling can
in depth and detail by capturing and terminated when no new information is
communicating the stories of the forthcoming from the new sampled. The
participants telling what happened, when, to approach used to collect data through open-
whom and with what consequences; they ended interviews involved outlining a list of
represent a creative synthesis of years of protocol questions prior to each interview,
participant observation and personal inquiry as a reminder regarding the information
(Patton, 2002).Open-end questions allows which needed to be collected, making sure
the respondent to express their thoughts that all relevant topics would be covered
freely, spontaneously, and in their own with each respondent and also making sure
language instead of forced to adapt to that the same basic lines of inquiry are
preconceived answers (Frankfort-Nachmias pursued with each person interviewed
and Nachmias, 1996), whereas documents (Patton, 2002; Yin, 2009).
are important to provide specific details, and Documents were used to corroborate and
to corroborate and augment evidence from augment evidence from other sources.
other sources (Yin, 2009). Relevant documents included newspaper
Focused interviews were carried out in articles, law decrees, reports from
Tegucigalpa from April 12 to May 9, 2010 CONASA, ERSAPS and different
with the actors identified as involved in the consultant companies, such as the Societe
actual decision-making process in the water Grenobloise d'Etudes et d'Application
sector: SANAA, CONASA, and ERSAPS. Hydrauliques (SOGREAH) and Pacific
Additional interviews were performed with a Consultants International (PCI), work from
consultant of SANAA, to two persons in different researchers and on-line reports
charge of water treatment plants and from the World Bank and IDB.
reservoirs, two SANAA trade union leaders Finally, case study evidence was examined to
(Sindicato Nacional de Trabajadores del discover causal links and build explanations
SANAA y Similares, SITRASANAAYS) and (Yin, 2009).
with the director of AMITIGRA, a NGO
working for the environmental protection of R ESULTS AND D ISCUSSION
the La Tigra national park, the site of the A false perception exists that the supply of
springs that feed the Picacho water plant. drinking water in the humid tropic is more
Due to the organizational complexity of than adequate to meet the needs of the
SANAA, the snowball sampling technique region, when in reality, the distribution of
was used as a method to identify rainfall is highly variable in time and space
interviewees. This technique consists of (Bonell et al, 1993). The Central America
locating one or more key individuals and region, for example, is characterized by a dry
asking them to name others who would be and rainy season and rainfall distribution is
likely candidates for the research (Russell, influenced the orientation of mountain
2000).The snowball sampling technique ranges and the configuration of coastlines
resulted in seven interviews with managers relative to the seasonal flow patterns
and two employees of different units of (Hastenrath, 1966). The alternating water
SANAA. The snowball sampling technique scarcity and abundance affects agriculture,
resulted in the sequence of the interviews human health and also the supply. A
shown in Figure 10.The first person selected sustainable management of water resources
to be interviewed using this technique was requires an inventory of what exists in the
the Sub manager of the National Water system, identification of the key processes,
Company (SANAA), Engineer Ricardo analysis of past environmental damage and
Velasquez due to his high influence in the an assessment of the capability of the
decision making process of SANAA. resource to support the various uses (Cullen,
The total of seventeen interviews were 1990).
considered to be sufficient because,
16
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
Fig. 11: Mean annual precipitation in the Choluteca River Catchment (millimetre) 1975–
1985 (upper row, for the 28 stations with more than 50% complete years with monthly
data during that period) and 1990–2005 (bottom row, for the 34 stations with more than
50% complete years with monthly data during the period) interpolated with inverse
distance weighting (left), universal kriging (middle) and ordinary kriging (right).
17
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
The mean monthly precipitation for the advantageous to use kriging instead of IDW
period 1990–2005 was higher in the upper in such cases. The low spatial
parts of the basin in the dry season autocorrelation for monthly precipitation in
December–April, with the opposite situation the dry season was likely associated with
in the rainy season May–November precipitation generated on windward sides
(Fig. 12). The analysis confirms the high of the mountains during northerly and
spatial variability of the precipitation created easterly winds.
by the precipitation-generating mechanisms. The high inter-annual precipitation
The mean cross-validation errors were in variability, with scattered temporal coverage
general lower for the UK and IDW methods of data, complicated the process of
for spatial interpolation of monthly time achieving a representative mean annual
series in 1970–2005. The difference was distribution. The interpolated mean annual
most clearly seen in the correlation precipitation regime was very different for
coefficient (Table 7). IDW was a better 1975–85 compared to 1995–2005, and it can
interpolator than the UK in the dry season be concluded that neither dataset gives a
when spatial dependence was low most of sufficient description of the spatial
the time. Goovarts (2000) states it is less precipitation regime. In the first period there
were more stations close to the coast, and in
Table 7: Cross-validation results for the later period there were few stations there
spatial interpolation of monthly but more in the upper catchment. However,
precipitation data from 1970–2005 even with more data available in the
with Ordinary Kriging (OK), mountainous upper areas in the later period,
Universal Kriging (UK) and Inverse the spatial variability is most likely highly
Distance Weighting (IDW). Errors underestimated. The ―bull‘s eyes‖ around
calculated as observed minus the mountain stations illustrate this problem,
if there were more stations in these areas the
interpolated.
high spatial variability would be more
Error measures 1970–
2005
OK UK IDW realistically described. A denser and
improved monitoring network is needed in
Correlation coefficient 0.26 0.40 0.43
order to achieve a sufficiently accurate
Mean absolute error (mm) 34 32 32 description of the local precipitation climate
Mean error (mm) 0.66 0.86 2.46 (especially in the mountains) for local
Max. positive error (mm) 134 126 133 applications.
Min. negative error (mm) -73 -78 -71 Once the precipitation has been quantified,
Mean relative error (%) 1.5 1.3 1.2 an approximation of the availability of water
Root mean square error
can be done. The estimated average annual
45 43 44
(mm)
18
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
precipitation in the Upper Choluteca Tatumbla River, the Guacerique and del
subcatchment is approximately 1,100mm per Hombre Rivers are better for the
year. Considering a total area of 816 km2 development of future water projects as they
with water supply potential, then this are larger and receive more precipitation.
precipitation represents a total of However, the availability of water is not
2,459,178m3/day. However, not all of the continuous through the year. The seasonality
precipitation will be transformed into of rainfall produces floods during the
surface runoff, which forms the surface months with high precipitation and droughts
storage. A part will evaporate and another during the rest of the year. This seasonality
part will infiltrate to form the ground water leaves water planners with a challenge of
storage which with time, will contribute to how to manage the available resources
the surface storage at a much slower rate efficiently and effectively.
(Chow et al, 1988). For this subcatchment,
Investment for a Sufficient Water
approximately 24.6% of the rainfall will be
transformed into surface runoff (Paz Supply (Paper II)
Barahona, 2003) leaving 604,957m3/day to It is because of the pronounced seasonality
satisfy the different needs of water. that the sources of water for Tegucigalpa
This research agrees with the conclusions partially dry up during the dry season. As a
drawn by SOGREAH and PCI that result the production of treated water drops
compared to Sabacuante River and during the dry season by 17% in
Fig 12: Mean monthly precipitation (millimetre) interpolated with IDW for stations
with more than 50% complete years of monthly data in the period 1990–2005. January
through April in the upper row (left to right), May through August in the second row
(left to right) and September through December in the bottom row (left to right).
19
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
20
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
It is very important to distinguish between patronage. In a spoils system, public jobs are
the cause and effects of the financial used as rewards for political activity (Simon
limitations. Assuming that SANAA would et al, 1991). The effects such a spoils system
be a financially healthy agency, then, the has had on SANAA are: (i) financial
Metropolitan division would have no need weakening due to the recruitment of
to subsidize other centres, it would have unnecessary personnel; (ii) lack of sense of
enough money to maintain and expand its identity between employees who realize their
network, reduce the unaccounted for water, job would be threatened in the next electoral
and would not be dependent on financial period; (iii) the lost of trained and highly
aid. Unfortunately, this is not the case. The qualified personnel for political reasons
origin of the financial limitations can (Flambard, 2003).
therefore be stated as the cost of the service Patronage has had another dire effect on the
and the amount of staff. water supply of Tegucigalpa which is the
The low price of the water service in lack of a long term planning strategy. A new
Tegucigalpa has been addressed by several manager and sub managers for SANAA are
researchers, who have recognized that appointed after each presidential election.
SANAA cannot expand or maintain the The frequent change of the senior
distribution network (Lee, 1996; Strand, management constitutes a serious obstacle
2000; Walker et al, 2000). SANAA has never to the continuation that should exist in
had control over the price which has been planning and actions, for it is not possible to
traditionally established on the basis of the execute long term plans and projects when
political interests, subject only to the the new administration wants to carry out
approval from the National Government their own (Caballero, 1992). It was pointed
(Strand, 2000). During the interviews it was out during the course of this research that a
mentioned that an increase of water tariffs is previous manager of SANAA had
only authorised when SANAA is unable to transformed it into a financially healthy
pay its employees. agency without increasing the price of water,
The high administrative costs originated but by promoting decentralization and
from the excess of employees has also been reducing the amount of staff
analysed using the efficiency indicator of (SITRASANAAYS, 1998). According to the
staff per 1000 connections and also by interviews, the successors continued the
observing the average annual cost of the tradition of recruiting political supporters, so
labour force per connection. In developing returning SANAA to its previous state of
countries the best practice target of staff per financial weakness.
1000 connections is 5 or less (Tynan and Because of the lack of financial resources
Kingdom, 2002). A high indicator is a result SANAA, has not been able to maintain the
of the political interference in the water existing network. The lack of maintenance is
company‘s operation (SITRASANAAYS, reflected in the high unaccounted for water
1998; Walker et al, 2000; Tynan and (UFW). UFW is defined as the difference
Kingdom, 2002). In the case of Tegucigalpa, between the amount of water pumped to the
Walker et al (2000) estimated that in 1994, system and the water metered to customers
this indicator was in the order of 13.6 and (Koch, 2010). High levels of UFW reflect
the annual cost of labour per connection the failure to repair leaks and meters, replace
was almost twice as much the cost of the old pipes, detect illegal connections,
connections in the rest of the country. In implement good commercial practices, and
2003 the employees of SANAA at a national meter all uses (Yepes, 1992). The two critical
level absorbed 70% of the utility budget and sides of the UFW are the financial burden
60% of those employees labour in for SANAA, who is producing water who
Tegucigalpa (CONASA, 2006). The origin no one is paying for, and also the loss of
of this excess of staff can be attributed to water which is much needed by the
what is known as the spoils system or population of Tegucigalpa and which is
21
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
probably just being wasted. The poor where Tegucigalpa lies by analyzing the
coverage of measuring devices throughout precipitation distribution and another to
Tegucigalpa makes it impossible to know seek for the reason why there has been a
how much water is lost. Studies have lack of investment in water supply projects.
estimated that the percentage of water lost The first analysis faced strong limitations
ranges from 30% (SOGREAH, 2003) up to originating from the poor coordination
61% (Hermida, 2006). Financially speaking, among agencies collecting meteorological
the volume of non-billed water can range data and the low quality of the collected
from 46% during the months corresponding information. A denser and improved
to the rainy season, to negative values during monitoring network, especially in the
the months corresponding to the dry season mountains, is needed to achieve a sufficient
(Hermida, 2006). The reasons behind this accurate description of the local
phenomenon are the flat rates and the precipitation climate for local applications.
intermittent water supply. During the dry However, the research has concluded that
season SANAA would reduce the number the average annual precipitation for the
of hours of service and the clients paying for stations in the Upper Choluteca
an average consumption would still have to subcatchment is approximately 1,100mm per
pay for a fixed volume of water that would year, with strong seasonal variability.
not be received (Hermida, 2006) and also This research coincides with the conclusions
because leakages do not exist if there is no drawn by SOGREAH and PCI regarding
water in the pipes. During the rainy season the feasible areas with potential for water
the number of hours of service per day is production projects. The areas with highest
increased, therefore the consumption is precipitation correspond to the Guacerique
higher and so is the volume that will be lost and del Hombre Rivers subcatchmnets.
in leakages. Because these areas as bigger as well, they
Either because of the financial limitations are more attractive for investment in water
originated from the imposed low price of supply structures from the hydrologic point
water, or from the excess of staff, the of view
Central Government has played a key role in
the financial problems of SANAA. The The precipitation gives an idea of the
financial limitations have lead to a lack of potential availability of water resources. But,
investment in expanding and maintaining how much water is needed to satisfy the
the distribution network which have needs of the population of Tegucigalpa?
furthermore lead to the actual water deficit Further research needs to be done to assess
in Tegucigalpa. whether this amount is sufficient or not. The
sufficiency will be determined by the
C ONCLUSIONS AN D F UTURE different water demands in the catchment.
R ESEARCH Once the different demands have been
quantified, a model can be calibrated. This
When Seckler et al (1999) analysed the water model would compare water demand and
supply and demand for 118 countries, they availability in an integrated and sustainable
mentioned that countries will face water way. This final step will provide the answer
scarcity by 2025, either because of the low to the question how much water is needed
precipitation, or because of the lack of to satisfy the needs in the subcatchment.
investment in water development projects.
In the case of Tegucigalpa, water scarcity is a The reasons causing the lack of investment
problem of the present. This thesis sought into developing water supply infrastructure
to answer the question of why Tegucigalpa were also analyzed. The analysis revealed the
does not possess a continuous water supply origins of the financial limitations were
by applying Seckler‘s explanations. So, two closely linked to the policies of Central
analyses were performed; one to quantify the Government. The Central Government has
available water resources in the catchment imposed low water tariffs on the water
22
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
23
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
R EFERENCES
Balairon Perez, L., Alvarez Rodriguez, J., Borrell Brito, E. and Delgado Sanchez, M. (2004).
Balance Hídrico de Honduras- Documento Principal. Madrid, CEDEX. 79p.
Berrios, F. (2009). Abandono de Proyecto Epas Agudiza Crisis de Agua. El Heraldo. Tegucigalpa.
34p.
Bonell, M., Hufschmidt, M. and Gladwell, J., Eds. (1993). Hydrology and Water Management in
the Humid Tropics. Hydrological Research Issues and Strategies for Water Management.
Cambridge, Press Syndicate of the University of Cambridge. 612p.
Caballero, E. (1992). Gestion Urbana y Participación Popular en la Ciudad de Tegucigalpa en la
Década de los 80. Tegucigalpa, Ediciones Zas. 34p.
Chow, V. T., Maidment, D. and Mays, L. (1988). Applied Hydrology. New York, McGraw-Hill.
572p.
Coello-Balthasar, Z., Phumpiu, P., Balfors, B. and Gustafsson, J. E. (2011). Assessment of Causes
Leading to an Insufficient Water Supply in Tegucigalpa, Honduras. Water Resources
Management VI, California, USA, WIT Press: 27-38
Cohen, B. (2004). Urban growth in developing countries: A review of current trends and a
caution regarding existing forecasts. World Development 32(1): 23-51.
CONASA (2005). Programa Plan Maestro de Abastecimiento de Agua para Tegucigalpa. Corto,
Mediano y Largo Plazo. Tegucigalpa, CONASA, DIAT. 27p.
CONASA (2006). Plan Estratégico de Modernización del Sector Agua Potable y Saneamiento
(PEMAPS). Tegucigalpa. 39p.
Cullen, P. (1990). The Turbulent Boundary Between Water Science and Water Management.
Freshwater Biology 24(1): 201-209.
Deal, J., Nazar, S., Delaney, R., Sorum, M., Leatherman, T., Gijsbert, N. F., Greene, G. and Wolf,
P. (2010). A Multidimensional Measure of Diarrheal Disease Load Charges Resulting
from Access to Improved Water Sources in Honduras. Practicing Anthropology 32(1):
15-20.
DGEC (1980). Primer Centenario de la Dirección General de Estadística y Censos Junio 1880-
1980. Tegucigalpa, Direccion General de Estadística y Censos: 115-117
DGEC (1989). Censo Nacional de Población y Vivienda 1988. Población Total y Numero de
Viviendas por Departamento y Municipio, Resultados Definitivos. Tegucigalpa,
Dirección General de Estadísticas y Censos: 22-75
Diaz Chavez, O. (1984). Estudio Hidrológico de Caudales Extremos para la Cuenca del Rio
Choluteca en Puente Choluteca. Facultad de Ingenieria. Tegucigalpa, Universidad
Nacionla Autonoma de Honduras. 189p.
Flambard, O. (2003). Rapport Actividad 1—Inventario, análisis y diagnostico de la red
hidrometeorologica, Electricité de France. ref D4168/RAP/2003-00059-B. 60p.
Frankfort-Nachmias, C. and Nachmias, D. (1996). Research Methods in the Social Sciences. New
York, Oxford University Press Inc. 592p.
Gleick, P. H. (1996). Basic water requirements for human activities: Meeting basic needs. Water
International 21(2): 83-92.
Goovaerts, P. (2000). Geostatistical approaches for incorporating elevation into the spatial
interpolation of rainfall. Journal of Hydrology 228(1-2): 113-129.
Hastenrath, S. L. (1966). Rainfall Distribution and Regime in Central America. Theoretical and
Applied Climatology 15(3): 201-241.
Heathcote, I. (1998). Integrated Watershed Management. New York, John Wiley & Sons, Inc.
414p.
24
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
25
Zairis Coello Midence Balthasar TRITA LWR PhD Thesis 2058
26
Insufficient water supply in an urban area – case study: Tegucigalpa, Honduras
Yepes, G. (1992). Infrastructure Maintenance in LAC: The Costs of Neglect and Options for
Improvement. G. Yepes, World Bank. 3. 29p.
Yin, R. (2009). Case Study Research: Design and Methods. California, SAGE Publications. 181p.
Other references
INE (2007). Trigésima Quinta Encuesta Permanente de Hogares Septiembre 2007 EPHPM -
XXXV. Honduras, Instituto Nacional de Estadística de Honduras. http://www.ine-
hn.org/-EPHPM_XXXV/survey0/data/Cuadros/02.%20Poblacion.mht. Accessed July
2010.
Morales, A. and Garcia, N. (2010). Ahora Desperdician el Agua. La Prensa. San Pedro Sula.
http://archivo.laprensa.hn/Pa%C3%ADs/Ediciones/2010/05/25/Noticias/Ahora-
desperdician-el-agua. Accessed May 2010.
Zepeda, F. (2006). Programacion de Producciones para el Año 2006 Tegucigalpa, SANAA. Electronic
spreadsheet.
27