1
GQ07: Securing Groundwater Quality in Urban and Industrial Environments (Proc. 6th International Groundwater
Quality Conference held in Fremantle, Western Australia, 2–7 December 2007).
Post-Audit of the State of Hawai`i’s Source Water
Assessment Program
S. R. SPENGLER1 & W. FREEMAN2
1 Pacific Hydrogeologic, 471 Iliwhai Loop, Kailua, Hawaii 96734
steve.spengler@gmail.com
2 Interactive Geographics
Abstract The US 1996 Safe Drinking Water Act Amendments mandated that
each state submit a source water assessment program (SWAP) that: (i)
delineates the boundaries of areas providing source waters for public water
systems, and (ii) identifies the origins of regulated and unregulated
contaminants in the delineated area to determine the susceptibility of public
water systems to such contaminants. The U.S. Geological Survey (USGS)
model MODFLOW was used to determine the source capture areas for the 104
public drinking water wells located on the island of Oahu, Hawai`i. Two
capture areas were defined using MODFLOW and the particle tracking code,
MODPATH: a 10-year travel time zone (Zone C), and a 2-year travel time
zone (Zone B). In addition to these two zones, a third area (Zone A), the “well
site control zone” with a diameter of 50 meters (m) around each source, was
established. Together, the three zones are referred to as Capture Zone
Delineations (CZDs). Once the CZDs were established, all potentially
contaminating activities (PCAs) within these zones were identified. Each PCA
was assigned a score (medium, high or very high), depending on the PCAs
relative potential to contaminate the underlying source water. Finally, each
water source was assigned a score based on the cumulative scores of the PCAs
identified within that source’s CZD. In this paper, an analysis is performed
that compares the susceptibility scores calculated for the groundwater supply
wells on Oahu to the actual occurrence of contamination in these wells.
Suggestions are made on how to improve the predictive capability of the
susceptibility methodology employed in the Hawai`i SWAP.
Key words source water assessment and protection; contaminated groundwater; capture zone
delineation
INTRODUCTION
Oahu is the most populated island in the State of Hawai`i and home to the State capital
and largest city, Honolulu. The resident and tourist population is about 950 000 people
and because much of the 586 square mile (1518 km2) island is mountainous and
covered by protected forest reserves, developed areas on Oahu are some of the most
densely populated areas in the United States. Urban and industrial development on the
island has been historically concentrated in Honolulu, on the coastal plain near Pearl
Harbor, and at several military bases. However, the central and south-western portions
of the island are currently in the process of undergoing major land use changes as a
result of the cessation and replacement of large-scale plantation agriculture (sugarcane
and pineapple) to suburban and diversified-crop agricultural use (Oki & Brasher,
2003).
2
S. R. Spengler et al.
The Honolulu Board of Water Supply (BWS) currently operates and maintains 104
water sources that combine to deliver an average of 155 million gallons of water per
day. More than 90 percent of this water is obtained from the deep volcanic rock
aquifers located in central Oahu and Honolulu, which has led to these systems being
designated as Sole Source Drinking-Water Aquifers by the United States
Environmental Protection Agency (USEPA). These basal aquifers are highly
permeable and unconfined except near the coast and in urban Honolulu, where they are
overlain by sedimentary confining units (Figure 1, from Oki et al., 1999). The
unconfined portions of these aquifers are susceptible to contamination resulting from
the downward migration of chemicals applied or spilled at the land surface. In central
Oahu, it has been estimated that contaminants can reach the basal water table within a
few years and persist in the aquifer and unsaturated zone for several decades (Orr and
Lau, 1988; Hunt, 2004). Vertical sampling of the basal aquifer in central Oahu has
revealed the stratified nature of the basal lens (Voss and Wood, 1994). The uppermost
water layer within the lens is 75 m to 125 m thick and consists of water recharged from
local rainfall and irrigation over the past few decades. Below this upper layer is the
core of the freshwater lens, 100 m to 150 m thick, containing waters with an apparent
carbon-14 age of 1,800 years, which floats atop the third layer composed of saltwater
that likely extends to the bottom of the aquifer. Groundwater ages were determined in
these aquifer systems during the National Water-Quality Assessment (NAQWA)
Program using chlorofluorocarbon and sulfur hexafluoride concentration data (Hunt,
2004). The NAQWA study found that the majority of groundwater samples collected
from Central Oahu and Honolulu had apparent ages ranging from the 1950s to 1980s,
which coincides with the period of the highest chemical use in agricultural chemicals
(predominately fumigants and herbicides) and urban settings (predominately
insecticides) (Anthony et al., 2004).
Fig. 1 Groundwater movement in Central Oahu and Honolulu aquifer systems.
Of the 90 NAQWA studies performed on major aquifer systems located
throughout the United States, the detection frequencies of fumigants, solvents,
insecticides and herbicides in Oahu public-supply wells were 1st, 3rd, 12th and 51st
nationwide, respectively (Anthony et al., 2004). The high frequency of pesticide
detection is related to the high historic rates of pesticide usage in agricultural settings
and the widespread use of solvents at military and civilian installations on Oahu. For
instance, it has been estimated that pesticide usage in Hawai`i was roughly 10 times
greater per square mile than amounts being used on the mainland during the 1960s
(State of Hawaii, 1969). Historically, fumigants and herbicides accounted for the vast
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
3
Post-Audit of the State of Hawai`i’s Source Water Assessment Program
majority of all pesticides consumed in Hawai`i. The fumigants were predominately
used by the pineapple industry (~85% of total state wide usage) for nematode control
while the sugar cane industry was the major single user of herbicides in the state,
accounting for about 75 percent of total herbicide usage in the State (State of Hawaii,
1969). Termite control operators (particularly in urban areas) are typically the greatest
users of persistent chlorinated hydrocarbon insecticides. In the 1960s, dosages
recommended for termite control in Hawai`i, up to 1500 pounds per acre, were roughly
500 to 1000 times higher per application than the amounts applied to agricultural areas.
Historically, chlordane was the most widely used persistent chlorinated hydrocarbon
insecticide. The mass of selected pesticides used throughout the State of Hawai`i in
1968 is summarized in Table 1.
Table 1 Total pesticide usage in State of Hawai`i in 1968.
Fumigants
[60.7%]
DD (1,2Dichloropropane)
EDB (1,2
dibromomethane)
DBCP (1,2Dibromo-3Chloropropane)
Other Chemicals
Total Statewide
Usage (Pounds
Active)
Pounds
Herbicide
[31.5%]
Pounds
Insecticides
[5.5%]
Pounds
6,050,000
Diuron
745,000
Chlordane
140,000
605,000
Dalapon
695,000
DDT
550,000
159,500
Atrazine
Other
380,000
1,998,900
Dieldrin
Other
7,364,500
3,818,900
Fungicides
[1.7%]
Pounds
58,000
116,000
Dithanes
Copper
Sulfate
16,000
398,000
Sulfur
Other
50,000
46,370
670,000
56,000
210,370
Fig. 2 Variation in pesticide and solvent concentrations in Central Oahu and Honolulu
aquifer systems over the past 25 years.
It is believed that the current contamination present in Oahu’s water supply wells
reflects historical use of chemicals more than present use (Anthony et al., 2004). The
presence of numerous agricultural and pest control chemicals such as TCP (1,2,3Trichloropropane), DPCP, and Dieldrin in the groundwater aquifers decades after their
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
4
S. R. Spengler et al.
use on Oahu was discontinued is evidence of the lasting impact of former pesticide
usage practices. The movement of groundwater in these basal aquifers is relatively
slow (typically less than 10 feet per day (≈ 3 m day-1)), and many of these chemicals
will likely take decades to be flushed through the aquifer systems. The State of
Hawai`i Department of Health (DOH) and the BWS currently test for more than 100
different types of contaminants on an annual basis. Figure 2 shows the concentration
level of fumigant, solvent, insecticide and herbicide contamination measured by DOH
and BWS in various wells located in central Oahu and Honolulu over the past 25 years.
SOURCE WATER ASSESSMENT PROGRAM METHODOLOGY
The reauthorization of the Safe Drinking Water Act (SDWA) in 1996 required states to
focus upon the protection of drinking water sources by development of a SWAP
(USEPA, 1997). The Safe Drinking Water Branch of DOH was the lead agency for
development of Hawai`i’s SWAP. The University of Hawai`i’s Water Resources
Research Center was contracted to conduct the assessments of Oahu’s 104 public
drinking water sources (Whittier et al., 2004). The assessment process consisted of the
following elements:
• Delineation of the area around a drinking water source through which
contaminants may travel to the drinking water supply;
• Inventory of activities that may lead to the release of microbiological or
chemical contaminants within the delineated area; and
• Determination of the susceptibility of the drinking water source to surrounding
potential contamination activities.
Delineation of capture zone
A total of three zones were delineated around each public groundwater supply system:
1) a well site control zone with a 50-meter diameter around each well site, which is
intended to assess the source’s vulnerability to tampering, vandalism and direct
introduction of contaminants (Zone A); 2) a two-year travel time zone to delineate the
area that may introduce pathogenic microorganisms directly into the water source
(Zone B); and 3) a ten-year travel time zone to delineate the area from which indirect
chemical contamination of a source could originate (Zone C).
The three-dimensional groundwater flow models, MODFLOW and MODPATH,
were used to determine the two- and ten-year capture zone around each source, based
upon the hydrogeology and pattern of groundwater withdrawal at each source (Whittier
et al., 2004). Two separate finite difference grid models were used to delineate the two
and ten-year capture zone delineations (CZD) for wells that serve public drinking
water systems on Oahu. The CZDs for wells located in Central and Leeward Oahu
were delineated using a model that covered all of Oahu except the eastern end of the
island (the Central and Leeward Oahu model). A second model, the Oahu model, was
extended to encompass the entire island of Oahu. The Central and Leeward Oahu
model consisted of 175000 cells (125643 active) arranged in 250 rows, 350 columns
and two layers which were rotated thirty degrees to better conform to the orientation of
the conceptual model. The Oahu model consisted of 283860 cells (203404 active)
arranged in 498 rows, 285 columns and two layers.
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
Post-Audit of the State of Hawai`i’s Source Water Assessment Program
5
Recharge from the USGS water budget prepared for Oahu based upon 1985-era
land use on the island was added to the model (Shade & Nichols, 1996). The pumping
rates used in the model were set at the allocated rates set by the State of Hawai`i
Commission on Water Resource Management. Calibration was performed by adjusting
the model parameters to make the model simulated hydraulic heads match the
measured hydraulic heads as closely as possible. The simulated water levels were
calibrated to the average water level for each well recorded in the 1999 USGS Water
Year report for Hawai`i (Hill et al., 2000). The primary parameter adjusted was
hydraulic conductivity, but adjustments were also made to conductance of the model
arcs representing streams, and to the hydraulic characteristic of the model arcs
representing volcanic dikes and streams. Using this approach, a very close calibration
was obtained for the basal water areas with the average error being less than 0.1 feet
(Whittier et al., 2004).
Inventory of potentially contaminating activities
Once the two and ten-year CZDs were determined using the MODFLOW model, all
PCAs were then inventoried within these two source water assessment areas. A PCA is
defined as a facility or activity that 1) stores, transmits, uses, or produces
contaminants, chemicals or by-products; and 2) has the potential to release
contaminants that may impact the quality of the underlying groundwater. A list of
PCAs specific to Hawai`i was created and ranked according to whether they posed a
very high (29 activities), high (23 activities) or medium (20 activities) contamination
potential to contaminate the source water. Factors that were considered when placing a
PCA into a certain category included 1) the nature of the activity, 2) chemicals
associated with the activity, and 3) association with historical incidents of groundwater
contamination. Table 2 provides examples of some of the more common PCAs that
were inventoried during development of Hawai`i’s SWAP.
Table 2 Examples of ranked potential contaminating activities inventoried.
Very High Contamination Potential
Large Quantity Hazardous Waste
Generators
Gas Stations
Dry Cleaners
Leaking underground storage tanks
(LUST)
Military Installation
High Contamination Potential
Automobile repair shops
Medium Contamination Potential
Above ground storage tanks
Diversified agriculture
Parking lots
Small Quantity Hazardous Waste
Generators
Septic systems
Parks
Residential parcels
Sewer Lines
Storm drain discharge points
Pineapple cultivation
Golf Courses
Storm water drainage - dry wells
Sugar cane cultivation
Underground storage tanks
Transportation corridors
The initial inventory of PCAs within the CZDs was created by querying existing
GIS data coverages of land use and hazardous waste release sites available from State,
County and private sources. All PCAs that fell within the delineated assessment areas
were integrated into a separate GIS layer created for the island of Oahu. The inventory
was largely confined to existing sites, with the exception of historic sites that had
already been inventoried and contained in DOH and USEPA hazardous waste
contamination databases (i.e. LUSTs, Resource Conservation and Recovery Act
(RCRA) sites, and Comprehensive Environmental Response, Compensation, and
Liability Act (CERCLA) sites). The location and existence of the PCAs identified
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
6
S. R. Spengler et al.
during the initial electronic inventory were then verified during a follow-on field
survey.
Susceptibility analysis
The susceptibility analysis determined the “relative potential” of a drinking water
source to be impacted by the PCAs that were identified in the delineated capture zone
during the inventory process. The purpose of the susceptibility analysis was to identify
drinking water sources that were most susceptible to contamination so that preventive
steps could be implemented. The scoring system shown in Figure 3 below was used to
determine a total susceptibility score for each groundwater source based upon the total
cumulative risk posed by the PCAs identified within that groundwater source’s source
water assessment protection area (i.e. the CZD). The total scores determined for
groundwater sources on Oahu ranged from 0 to 2,119, with the highest scores typically
being associated with wells located in urban settings.
Fig. 3 Scoring system to calculate groundwater source susceptibility to contamination
from PCA within delineated capture zone.
RESULTS
Figure 4 shows the two and ten-year CZD for public groundwater supply wells
calculated by MODPATH as well as the location of contaminated and noncontaminated wells on Oahu.
One of the objectives of the SWAP was to provide the data required for
development of a comprehensive source water protection strategy for the State of
Hawai`i to ultimately prevent contaminants from entering public water supply systems.
The susceptibility analysis methodology was developed to rank the relative
susceptibility of the groundwater wells to contamination by analyzing the types and
number of potentially contaminating activities located within the land area that overlies
the extent of groundwater that contributes water to the well over a ten year time period.
It is thus instructive to compare the susceptibility scores calculated during the SWAP
program with the actual occurrence of contamination within the water supply wells.
For the purposes of this comparison, we reduced the scores of well sources at which
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
Post-Audit of the State of Hawai`i’s Source Water Assessment Program
7
contamination had been detected by 50 to remove the bias towards contaminated wells
built into the scoring system (i.e. 50 points were added to wells in which
contamination has been detected in the original SWAP methodology). The correlation
between the total PCA score and the occurrence of groundwater contamination is
shown in Figure 5. In this figure, a single score was assigned to co-located wells which
had overlapping CZDs and thus identical PCA scores.
Fig. 4 Two and ten-year capture zones delineated on Oahu using MODPATH.
Fig. 5 Correlation between PCA score and occurrence of groundwater contamination.
DISCUSSION
Figure 5 shows that the susceptibility analysis used for the Hawai`i SWAP was not
particularly predictive for determining the occurrence of groundwater contamination.
The wells with very high PCA scores (>500) were typically located in urban
environments with high densities of PCAs while many of the wells located in Central
Oahu in the vicinity of former and current agricultural fields typically yielded total
GQ07: Securing Groundwater Quality in Urban and Industrial Environments
8
S. R. Spengler et al.
scores between 100 to 500. The predictive capability of the susceptibility analysis for
Oahu could be improved somewhat if two additional factors were accounted for in the
scoring analysis: 1) the mass of pesticide historically applied within the CZD and 2)
the hydrogeologic setting (i.e. whether the CZD overlies a confined or unconfined
aquifer). Figure 4 shows that virtually all of the contaminated groundwater sources on
Oahu are located in central Oahu or western/central Honolulu. The central Oahu wells
reflect residual contamination from the former large tracts of agricultural land in the
area (sugarcane and pineapple) which received repeated applications of fumigants and
herbicides over the past sixty years. These pesticides were applied over a vulnerable
unconfined aquifer. Solvents and petroleum products have also been used and stored at
several military installations in this area since the 1940s. The detection of these types
of contaminants in central Oahu likely reflects historic releases from these
installations, or other civilian sources, such as gasoline stations and automotive repair
shops. In contrast to central Oahu, few organic compounds were detected in Honolulu
wells despite the high urban density and the intensive use of insecticides in this area.
Trace levels of the insecticides dieldrin and chlordane were detected in the Aiea to
Kalihi areas of Honolulu. The absence of significant contamination in the wells located
in the Honolulu area (which typically have high PCA scores due to the numerous
PCAs within these wells CZDs) is due to these wells tapping confined aquifers, which
afford some protection from downward migration of surface contamination.
The susceptibility analysis could be improved by adopting a fourth contamination
potential category for current and former agricultural areas where the magnitude of the
scores assigned for these PCAs are a function of both the area within the CZD which
was in agriculture and the historic amount of pesticide applied to this area. In addition,
the total score should be adjusted to reflect whether a particular well’s CZD is located
over a confined (decrease score) or unconfined aquifer (increase score).
REFERENCES
Anthony, S. S., Hunt, Jr., C. D., Brasher, A. M., Miller, L. D. & Tomlinson, M. S. (2004) Water quality on the island of
Oahu, Hawaii. US Geol. Survey Circular 1239, 37 p.
Hill, B. R., Fontaine, R. A., Taogoshi, R. I. & Teeters, P. C. (2000) Water Resources Data: Hawaii and Other Pacific
Areas, Water Year 1999. Volume 1. Hawaii, 399 p.
Hunt, C. D., Jr. (2004) Ground-water quality and its relation to land use on Oahu, Hawaii, 2000–2001. US Geol. Survey
Water-Resources Investigations Report 03–4305, 76 p.
Oki, D. S., Gingerich, S. B. & Whitehead, R. L. (1999) Hawaii in Ground Water Atlas of the United States, Segment 13,
Alaska, Hawaii, Puerto Rico, and the U.S. Virgin Islands: US Geol. Survey Hydrologic Investigations Atlas 730-N, p.
N12–N22, N36.
Oki, D. S. & Brasher, A. M. D. (2003) Environmental setting and the effects of natural and human-related factors on water
quality and aquatic biota, Oahu, Hawaii. U.S. Geological Survey Water-Resources Investigations Report 03-4156,
98.
Orr, S. & Lau, L. S. (1988). Modeling of Trace Organic (DBCP) Transport in Pearl Harbor Aquifer, Oahu, Hawaii:
Method of Characteristics, Phase II. University of Hawaii, Water Resources Research Center Technical Report No.
175.
Shade, P. J. & Nichols, W. D. (1996) Water Budget and the Effects of Land-Use Changes on Ground-Water Recharge,
Oahu, Hawaii, U.S. Geol. Survey Prof. Paper 1412-C.
State of Hawaii, Department of Agriculture. (1969) Evaluation of Pesticide Problems in Hawaii. Report prepared by the
State of Hawaii, Department of Agriculture, Honolulu, Hawaii.
USEPA (1997) State Source Water Assessment and Protection Programs Guidance Draft Guidance, EPA Number
816R97007, 108 pp.
Whittier, R., Rotzoll, K., Dhal, S., El-Kadi, A., Ray, C., Chen, G. & Chang, D. (2004) Hawai`i source water assessment
program report, County of Honolulu, Hawai`i Department of Health.
Voss, C. I. & Wood, W. W. (1994) Synthesis of geochemical, isotopic and ground-water modeling analysis to explain
regional flow in a coastal aquifer of southern Oahu, Hawaii. In: Mathematical Models and their Applications to
Isotope Studies in Groundwater Hydrology, International Atomic Energy Agency (IAEA) Vienna, Austria, IAEATECDOC-777, 147-178.
GQ07: Securing Groundwater Quality in Urban and Industrial Environments