Among island communities, water security and access continue to be a critical issue. In the US Virgin Islands (USVI), 90% of households are capable of collecting roof-harvested rainwater, whereas less than 25% of households are connected to a municipal water system serviced by desalination plants. Rainwater collection leaves the household in charge of managing and treating their own water. Therefore, understanding current barriers to accessing water treatment technologies and the costs of a water storage and treatment program at scale are critical in developing a territory-wide program. In this study, we evaluate (1) household-level barriers to accessing water treatment technologies, (2) a cost breakdown of a UV water treatment pilot program, and (3) potential estimates of program costs at a territory-wide scale. The results suggest that for households, key barriers include knowledge of the problem or solution and capital and installation costs. From the programmatic cost evaluation, the water treatment technology and water testing services were the most expensive. However, given key adjustments, a territory-wide program is estimated to cost $30.4 M covering 50% of households. These data can be used by a coalition of stakeholders in allocating financial and managerial responsibility for a territory-wide water storage and treatment program in USVI.

  • Households in USVI are responsible for treating and storing their water supply.

  • The primary barriers to acquiring adequate treatment are knowledge of the problem and capital costs of the technology.

  • A water treatment pilot program identified the highest cost components as the treatment technology and water testing.

  • A territory-wide water treatment program is estimated to cost $30.4 M to cover 50% of the population.

Access to potable water in the US Virgin Islands (USVI) continues to be a challenge for vulnerable populations while the lack of critical infrastructure and the effects of climate change exacerbate the situation. Therefore, it is critical to develop systematic and strategic approaches to climate resilience for water access among households. As of 2019, greater than 90% of USVI households have the capability of collecting roof-harvested rainwater in large cisterns, 25% of households are connected to the municipal water system (10–15% have both options), and less than 1% use private well water (Government of the United States Virgin Islands, 2018). Typical rainwater harvesting systems include a roof catchment system with downspouts directed into below-ground concrete cisterns and a pump-pressure tank system to distribute cistern water throughout the household. Households connected to the municipal water system either receive water directly into their premise plumbing system (pressurized by the municipal water system) or receive water into their cistern which is subsequently distributed throughout the household via the pump-pressure tank system. Water sources used for daily household activities may include cistern water, municipal water, or bottled water (Government of the United States Virgin Islands, 2018; Rao et al., 2022).

If cistern water is available, it is used for the majority of household activities including cooking, cleaning, washing fruits and vegetables, showering and bathing, and brushing teeth (Rao et al., 2022). Furthermore, in a recent evaluation of water use habits among households in USVI, 21% used cistern water as their primary source of drinking water, while 40% of households used cistern water as either a primary or secondary drinking water source (Rao et al., 2022). However, Escherichia coli was detected in 80% of cistern water samples (n = 371), and at least one type of waterborne pathogen was detected in 66% of cistern water samples (Rao et al., 2022). In a 2019 review of rainwater harvesting systems (n = 17 studies), E. coli was detected in 54% of all cistern water samples reported globally (n = 1,117 samples) (Hamilton et al., 2019). These data suggest that, if available, directly piped municipal water may be the best option, while if roof-harvested cistern water is used, treatment may be necessary depending on the type of domestic water use activities, especially drinking. Furthermore, if the same cistern used for storing rainwater is also used for storing municipal water, this water is also at risk of contamination and needs to be treated. However, due to the decentralized and private nature of the water source, the primary responsibility falls on the household for (1) minimizing the risk of contamination, (2) deciding if water treatment is necessary for the types of household water activities conducted daily, (3) if necessary, what type of water treatment option is optimal and how to acquire it, and (4) how to operate the water treatment system in a sustainable and effective manner. To support households in these responsibilities, external stakeholders have a variety of options given the type of entity (e.g., non-governmental organization (NGO), local government, and small business), available funding, and status in the community.

Like technology, developing any program, service, or policy must have the end-user or beneficiary central to the design and development process. For improving access to potable drinking water among USVI households, the household and the problems they face with water storage and treatment options are the central focus. Household problems may include awareness of the problem and/or solutions; the cost of the contamination prevention measure or treatment system; the cost and ability to install the treatment system; the cost, access to, and timely replacement of system parts; training on the operation and maintenance of the treatment system; testing to ensure system effectiveness; and troubleshooting, diagnostics, and corrective actions for a failed or broken treatment system. A program, service, or policy does not only need to address all problems household's face but will also depend on (1) the primary barriers for households to appropriately use effective water treatment options, (2) the type of external stakeholder wanting to engage, and (3) the resources available to the stakeholder. The previous work has demonstrated approaches addressing some of these problems via rebate models for rain harvesting (Imteaz & Moniruzzaman, 2018), optimizing household and community water balance models (Imteaz & Boulomytis, 2022), and improving roof and tank sizing capabilities (Shadeed & Alawna, 2021).

For locally available water treatment options, previous work has been conducted to evaluate or predict the effectiveness of specific water treatment technologies including chlorination (Voth-Gaeddert & Schranck, 2021; Lindmark et al., 2022; Voth-Gaeddert et al., 2022a, 2022b), filtration (Voth-Gaeddert et al., 2022a), and UV (Voth-Gaeddert et al., 2022c) at the levels of whole-house (point-of-entry) and point-of-use. In addition, mechanisms to protect household cistern water from becoming contaminated have also been identified (Rao et al., 2022). Table 1 summarizes these findings.

Table 1

Water storage practices and water treatment options.

Device/approachDescriptionCitation
Water storage practices 
Cover overflow drains or other entryways to the cistern All cisterns have at least one overflow pipe. Recommended practice is to securely attach a wire mesh over the opening to mitigate animal or other intrusion. Rao et al. (2022)  
Well-sealed cistern access points Poorly fitted or open cistern hatches or access doors can allow for animals, debris, or contaminated liquid to enter the cistern. Rao et al. (2022)  
Reducing the presence of ground animals near the cistern Observing ground animals such as iguanas, rodents, frogs, or similar near the cistern can be an indication they can get in. Following guidelines around pest management (e.g., limiting debris piles) can help reduce presence. Rao et al. (2022)  
Mosquito presence and control issues Cisterns can be an ideal breeding ground of mosquitos; therefore, ensuring proper management and storage is important. Seger et al. (2022)  
Branches overhanging the roof Branches overhanging roofs may provide opportunities for animals to access the roof or directly defecate onto the roof. Rao et al. (2022)  
First flush systems on downspouts First flush systems are often installed on downspouts and divert the first amount of water away from the cistern during a rain event as it is often the most contaminated and contains debris. Charlebois (2021)  
Water treatment systems 
UV systems with prefilters Point-of-entry, installed after pump and pressure tank; prefilters help reduce turbidity; UV system should be rated to handle the maximum flow rate in household (e.g., 6–8 gallons/min) and provide sufficient UV fluence (dose). Voth-Gaeddert et al. (2022c)  
UV systems without prefilters Same install as the prior UV system, but if cost savings are substantial, prefilters may be omitted if turbidity levels are low (<5–8 NTUs). Voth-Gaeddert et al. (2022c)  
Passive chlorination (in developmentPoint-of-entry, installed after pump and pressure tank; chlorine tablets are most common (NSF/ASNI 60 certified); dosing level and contact time must be high enough to produce sufficient CT values; a 1-μm absolute membrane filter is needed for parasite removal while a carbon filter can be added prior to the tap to remove unpleasant chlorine tastes or odors. Voth-Gaeddert et al. (2022a) and Lindmark et al. (2022)  
Under-the-sink treatment options Options include RO systems which can be directly tied into sink plumbing; mini-UV systems are also available and can be smaller as the flow rate from a single tap is lower. Voth-Gaeddert et al. (2022c)  
Counter-top treatment options These treatment options sit on top of the counter and can include small RO systems often tied into the sink plumbing; ceramic filters are effective but stand-alone systems where water is added into the top compartment and slowly percolates through the filters to the bottom compartment. Berkey Filters (2022) a 
Batch chlorination (not effectiveA common local practice in USVI is pouring bleach into the cistern. This approach is very difficult to facilitate correctly and can lead to overdosing of the cistern. It is not recommended for treating microbially contaminated cistern water for drinking purposes. Voth-Gaeddert et al. (2022b)  
Device/approachDescriptionCitation
Water storage practices 
Cover overflow drains or other entryways to the cistern All cisterns have at least one overflow pipe. Recommended practice is to securely attach a wire mesh over the opening to mitigate animal or other intrusion. Rao et al. (2022)  
Well-sealed cistern access points Poorly fitted or open cistern hatches or access doors can allow for animals, debris, or contaminated liquid to enter the cistern. Rao et al. (2022)  
Reducing the presence of ground animals near the cistern Observing ground animals such as iguanas, rodents, frogs, or similar near the cistern can be an indication they can get in. Following guidelines around pest management (e.g., limiting debris piles) can help reduce presence. Rao et al. (2022)  
Mosquito presence and control issues Cisterns can be an ideal breeding ground of mosquitos; therefore, ensuring proper management and storage is important. Seger et al. (2022)  
Branches overhanging the roof Branches overhanging roofs may provide opportunities for animals to access the roof or directly defecate onto the roof. Rao et al. (2022)  
First flush systems on downspouts First flush systems are often installed on downspouts and divert the first amount of water away from the cistern during a rain event as it is often the most contaminated and contains debris. Charlebois (2021)  
Water treatment systems 
UV systems with prefilters Point-of-entry, installed after pump and pressure tank; prefilters help reduce turbidity; UV system should be rated to handle the maximum flow rate in household (e.g., 6–8 gallons/min) and provide sufficient UV fluence (dose). Voth-Gaeddert et al. (2022c)  
UV systems without prefilters Same install as the prior UV system, but if cost savings are substantial, prefilters may be omitted if turbidity levels are low (<5–8 NTUs). Voth-Gaeddert et al. (2022c)  
Passive chlorination (in developmentPoint-of-entry, installed after pump and pressure tank; chlorine tablets are most common (NSF/ASNI 60 certified); dosing level and contact time must be high enough to produce sufficient CT values; a 1-μm absolute membrane filter is needed for parasite removal while a carbon filter can be added prior to the tap to remove unpleasant chlorine tastes or odors. Voth-Gaeddert et al. (2022a) and Lindmark et al. (2022)  
Under-the-sink treatment options Options include RO systems which can be directly tied into sink plumbing; mini-UV systems are also available and can be smaller as the flow rate from a single tap is lower. Voth-Gaeddert et al. (2022c)  
Counter-top treatment options These treatment options sit on top of the counter and can include small RO systems often tied into the sink plumbing; ceramic filters are effective but stand-alone systems where water is added into the top compartment and slowly percolates through the filters to the bottom compartment. Berkey Filters (2022) a 
Batch chlorination (not effectiveA common local practice in USVI is pouring bleach into the cistern. This approach is very difficult to facilitate correctly and can lead to overdosing of the cistern. It is not recommended for treating microbially contaminated cistern water for drinking purposes. Voth-Gaeddert et al. (2022b)  

aHere, we provide this reference as only an example of the product and not an endorsement.

In this paper, we aim to leverage existing and new data to (1) identify critical components for a territory-wide program and (2) provide the relative costs of the core components necessary for the program. First, we evaluate secondary data from a territory-wide survey of households that included questions on access to treatment technologies and desired support services. Second, we evaluate a pilot program conducted in USVI by a local NGO which provided household access to water treatment technologies and services. Stakeholders can use these estimates to decide on the equitable allocation of financial and managerial responsibility for a territory-wide water treatment and storage program across the different components.

Setting

USVI has a population of 106,405 people and 40,648 households across three primary islands, such as St. Croix (pop: 50,601), St. John (pop: 4,170), and St. Thomas (pop: 51,634) (United States Census Bureau, 2010). 8.4% of the population is >65 years old, while 31.6% are <18 years old (United States Census Bureau, 2010). In addition, over 45% of people have some type of chronic health condition (Rao et al., 2022). As of 2010, the median income among USVI households is $24,704, while 32.5% of households live below the poverty line (United States Census Bureau, 2010). As of 2022, stores that may stock certain types of water treatment systems include local and national hardware stores (two each on St. Croix and St. Thomas, one on St. John) as well as pool supply stores (one each on St. Croix and St. Thomas). These systems could include whole-house (or point-of-entry) treatment systems such as a UV system with pre-filtration or an inline chlorinator with a 1-μm carbon filter or point-of-use systems such as counter-top ceramic filters or reverse osmosis (RO) systems. Finally, households that do not have access to the municipal water system use water tanker trucks to refill their cisterns when rainwater is insufficient. This water is provided by private third-party companies who obtain water from the municipal water utility (WAPA). This cost of water is expensive because it is largely dependent on energy prices because all municipal water is generated by desalination plants (owned and operated by a private third party).

Given the high capital costs and household distribution across the islands, expansion of the municipal water system is costly and difficult. In addition, the practice of rain harvesting will continue to be a primary method used in USVI and continue to rise in popularity across the Caribbean and beyond. Therefore, we present details on core components necessary for achieving territory-wide access to potable water in USVI. Here, we provide details on a territory-wide survey used to collect initial household preferences and an evaluation of a pilot program which provided households with access to UV water treatment systems. Figure 1 depicts the framework used to integrate the primary and secondary data sources for this paper. First, the detailed methods and primary results of the territory-wide survey conducted in 2019 are reported by Rao and colleagues; however, secondary data analysis was conducted on survey data collected in that study. Second, detailed methods and primary results of a UV water treatment pilot program conducted between 2019 and 2021 in St. John are reported by Voth-Gaeddert and colleagues; however, secondary data analysis was conducted on program costs. Finally, these results were then used to estimate the cost of a territory-wide water storage and treatment program for USVI households.
Fig. 1

Framework of integrating primary and secondary data for analysis. *See Table 1 and each publication for additional details.

Fig. 1

Framework of integrating primary and secondary data for analysis. *See Table 1 and each publication for additional details.

Close modal

USVI cistern survey evaluation

In July and August of 2019, the USVI Department of Health and the Centers for Disease Control and Prevention conducted a territory-wide survey of cistern water quality in the USVI. Rao et al. (2022) described the methods and primary results of the evaluation. As part of this evaluation, a household survey was conducted using a locally validated questionnaire and conducted verbally via a field team member with the head of the household. Households were selected via two-stage random sampling using census blocks and only those households who had cisterns were surveyed (see Rao et al., 2022 for full methods). As part of the survey (n = 399), two questions were asked about using multi-stage UV treatment systems with prefilters (locally known as ‘multi-stage water treatment systems’) and preferences for a water treatment access program. These questions included:

  • Multi-stage water treatment systems can potentially be an effective option to treat your water. However, there are barriers to obtaining these multi-stage systems. What are the biggest reasons for not getting a multi-stage water treatment system?

All responses given by the household were marked and any additional comments not available on the list were put in the ‘other’ category as a write-in.

  • Which of the following options would cause you to be more likely to getting a multi-stage water treatment system?

All response options were read to the interviewee, and they were allowed to select up to two options or write-in their own response. Options included: the cost of the system was less; the cost of installation was less; buying, installing, and starting the system was bundled; or there was a 1- to 2-year payment plan available for purchasing the system. Data reported in the Supplementary Material in Rao et al. (2022) were aggregated and basic descriptive statistics were generated. For text responses provided in the ‘other’ option, responses were grouped and any groups which aligned with available options were integrated back in (some persons selected ‘other’ but specified an available option).

UV water treatment pilot program

From 2019 to 2021, Love City Strong, Inc. (LCS), a local NGO, enrolled households from St. John in an access program for UV water treatment systems. The structure of the program and specific results of the effectiveness of the UV treatment systems themselves is reported in Voth-Gaeddert et al. (2022c). Here, we evaluated the total cost of the program via a cost-per-household metric as well as a cost breakdown for the different program components. These components included (i) the cost of the system; (ii) the cost and ability to install the system; (iii) the cost of replacement parts for the system; (iv) training on operating and maintaining the system; (v) testing to ensure system effectiveness; and (vi) troubleshooting, diagnostics, and corrective actions for failed or broken systems.

Briefly, for the program, 65 households were enrolled over the course of 2 years. The program provided households with a Rainfresh Rainwater Filtration [and UV] System (model #: RW8; Richmond Hill, Ontario, Canada) installed for free by a certified plumber directly after the pump-pressure tank system. Households were provided training on how the system worked, how to operate and maintain the system, and with a 1-year supply of replacement parts for the system. In addition, the household was visited by the field team every month for the first 6–12 months after enrollment and installation of the system. Upon each visit, the field team evaluated the operation of the system and collected water quality data to evaluate the effectiveness of the system (see Voth-Gaeddert et al., 2022c for details). A baseline and exit survey was conducted to collect data on household characteristics and water use habits of the households in the program. Finally, if the water quality testing indicated that the water quality post-treatment was contaminated (i.e., defined as any level of detection of E. coli via the IDEXX-Colilert-24® and the Quanti-Tray/2000 method (American Public Health Association (APHA), 2018)), and corrective action was taken by the field team.

As Voth-Gaeddert et al. (2022c) extensively evaluated the cost-effectiveness of the UV treatment devices from the perspective of the household, here we focused on the program costs from the perspective of the implementing stakeholders. We evaluated (1) What percent of households had at least one water sample with detectable levels of E. coli? (2) What percent of households reported using cistern water for additional household activities at the end of the program? and (3) The cost-per-household for the program in addition to a detailed cost breakdown of each component along with program optimization options. Each component listed above was evaluated for actual costs as of 2022. Both retail and wholesale costs are provided, where applicable. Finally, we discussed what these costs may be at scale (10% or 4,064 households and 50% or 20,324 households).

USVI cistern survey evaluation

To acquire an initial understanding of current local sentiment and knowledge toward water treatment systems, secondary data from Rao et al. (2022) were analyzed (see Tables 2 and 3). First, in response to the question, ‘What are the biggest reasons for not getting a multi-stage water treatment system?’, the top two most frequent responses were related to knowledge of the problem or solution, i.e., ‘did not consider…’ (31.1%) or ‘do not need…’ (17.5%). The next two most frequent responses included issues related to the costs of these systems: capital costs (14.4%) or installation costs (10.8%). While multiple responses were allowed and not all responses were mutually exclusive, these top two general issues of knowledge and cost captured a majority of those surveyed.

Table 2

Household responses to barriers to obtaining a multi-stage (prefilters + UV) household water treatment system (n = 360).

Response optionsCountPercent
The upfront cost is too high 52 14.4 
Did not consider a multi-stage system 112 31.1 
The time and money to buy and install the system is higher than the potential health benefits 18 5.0 
Installation cost is too high 39 10.8 
Do not need a multi-stage system 63 17.5 
Operation and maintenance cost is too high 28 7.8 
Operation and maintenance time, stress, and energy are too high 15 4.2 
Other (specify) 48 13.3 
N/A I'm a renter 34 9.4 
Too many steps are involved in buying, installing, and starting a system 22 6.1 
Response optionsCountPercent
The upfront cost is too high 52 14.4 
Did not consider a multi-stage system 112 31.1 
The time and money to buy and install the system is higher than the potential health benefits 18 5.0 
Installation cost is too high 39 10.8 
Do not need a multi-stage system 63 17.5 
Operation and maintenance cost is too high 28 7.8 
Operation and maintenance time, stress, and energy are too high 15 4.2 
Other (specify) 48 13.3 
N/A I'm a renter 34 9.4 
Too many steps are involved in buying, installing, and starting a system 22 6.1 

Note: Additional responses were less than 1.7% and are listed in the Supplementary Material; variable prefix in the Supplementary Material MSTS_Barriers; no response options were read aloud by the interviewer, but any were marked when mentioned.

Table 3

Household responses to which attributes would increase the likelihood of the household obtaining a multi-stage (prefilter + UV) water treatment system (n = 360).

Response optionsCountPercent (%)
Buying, installing, and starting the system were bundled into one 58 16.1 
The cost of installation was less 36 10.6 
The cost of the system was less 74 22.5 
Others, specify 84 23.3 
 Other write-in: need/want more info  29 8.1 
 Other write-in: not interested  17 4.7 
There was a 1- or 2-year payment plan for the system 22 6.1 
N/A, I'm a renter 36 10.0 
Response optionsCountPercent (%)
Buying, installing, and starting the system were bundled into one 58 16.1 
The cost of installation was less 36 10.6 
The cost of the system was less 74 22.5 
Others, specify 84 23.3 
 Other write-in: need/want more info  29 8.1 
 Other write-in: not interested  17 4.7 
There was a 1- or 2-year payment plan for the system 22 6.1 
N/A, I'm a renter 36 10.0 

Additional ‘other’ responses are listed in the Supplementary Material; all response options were read and the interviewee could select/write-in two.

Second, in response to the question, ‘Which of the following options would cause you to be more likely to getting a multi-stage water treatment system?’ the most frequent response was ‘other’ (23.3%) with a write-in of ‘need/want more info’ or ‘not interested’. The second most frequent response was that the initial or capital cost of the system was less (22.5%). This aligns with responses to the first question and provides a useful set of priorities, including information and education, for a territory-wide program.

UV water treatment pilot program

The pilot program included many of the core components often incorporated in similar programs which allowed for a full evaluation in addition to the ability to identify both the expensive and critical components of the program. Overall, for individual water samples (n = 271), 95.2% of tap samples had no detectable E. coli (Voth-Gaeddert et al., 2022c). At a household level (n = 65), eight (12%) had at least one post-treatment sample with detectable levels of E. coli over the 12-month duration of the free water testing program component (five had only one, two had two, and one had more than two samples with detectable levels of E. coli). In addition, as reported by Voth-Gaeddert and colleagues, only one household had positive detections of E. coli in water samples in back-to-back months. The authors note that this may suggest that support services were helpful (corrective action), but that clear operation and maintenance guidance must be provided to the household as ‘treatment failures’ may have been primarily associated with user errors (e.g., opening a bypass). Finally, of the n = 65 households in the program, only five (8%) reported the expanding use of cistern water for household water use activities beyond what they reported at baseline. This may reflect that households already use untreated cistern water for most daily water-based activities or that general cultural norms around drinking water are well established.

Table 4 presents the program cost breakdown and overall cost-per-household value that implementing stakeholders would incur. The total wholesale cost-per-household for the program LCS facilitated was $2,599. The most expensive components of the program were the UV treatment system and the water testing. It is important to note that these costs are subject to change based on a variety of factors including global market forces, the type of treatment system used (which influences the installation costs and replacement part costs), and how personnel time is allocated to various support services for households. In addition, there are potential cost savings in how a water testing program is developed and executed.

Table 4

Detailed cost breakdown for the program and cost-per-household estimate.

DescriptionCost – retailCost – wholesale
Cost of the UV treatment system   
 Cost of a UV system in the box $1,189 $582 
Cost of the installation   
 Additional parts (PVC, joints, etc.) $50 $50 
 Plumbing labor (∼$100/h) $350 $350 
Cost of replacement parts   
 UV bulb (annual replacement) $110 $44 
 Prefilters (×2, replace every 6 months) $212 $120 
Training on operating and maintaining the system (1 training visit)   
 Personnel time $30 $30 
Water testing for system effectiveness (12 visits; 24 samples)   
 Direct testing costs (reagents, equipment, etc.)a $213 $213 
 Personnel time (visits + lab processing time) $1,080 $1,080 
Troubleshooting, diagnostics, and corrective actions (1 visitb  
 Personnel time $30 $30 
 Plumbing labor $100 $100 
Total cost-per-household (n = 65)c $3,364 $2,599 
DescriptionCost – retailCost – wholesale
Cost of the UV treatment system   
 Cost of a UV system in the box $1,189 $582 
Cost of the installation   
 Additional parts (PVC, joints, etc.) $50 $50 
 Plumbing labor (∼$100/h) $350 $350 
Cost of replacement parts   
 UV bulb (annual replacement) $110 $44 
 Prefilters (×2, replace every 6 months) $212 $120 
Training on operating and maintaining the system (1 training visit)   
 Personnel time $30 $30 
Water testing for system effectiveness (12 visits; 24 samples)   
 Direct testing costs (reagents, equipment, etc.)a $213 $213 
 Personnel time (visits + lab processing time) $1,080 $1,080 
Troubleshooting, diagnostics, and corrective actions (1 visitb  
 Personnel time $30 $30 
 Plumbing labor $100 $100 
Total cost-per-household (n = 65)c $3,364 $2,599 

aWe ignored capital equipment costs (Quanti-Tray/2000, Incubator, autoclave, and black light reader), only including Colilert medium, Whirl-Pak bags, and trays totaling $8.88 per sample. Electricity costs were ignored as those are direct charges to households (see Voth-Gaeddert et al., 2022c for a detailed breakdown of direct charges to households).

bBased on Voth-Gaeddert et al. (2022c), less than one visit per household was needed for rectifying a positive E. coli test during the program.

cTotal costs assume the program facilitates engagement for 1 year and no programmatic costs are incurred after the first year (households are entirely in charge).

USVI cistern survey

In this study, we aimed to complement the previous work on water storage practices and treatment options by evaluating community perceptions of water treatment technologies and programs as well as evaluating the cost-per-household for a UV water treatment pilot program. The survey results were consistent in demonstrating three key needs to improving community members’ perceptions related to obtaining an effective water treatment system: (1) clear and accessible information on effective water treatment systems, (2) clear and accessible information on WHY a household should invest in a water treatment system, and (3) the costs of water treatment systems (or ‘business as usual’/opportunity costs) along with ways households can mitigate or lower costs to invest. Furthermore, the fact that few households expanded the use of cistern water to additional household activities within the UV program after access to water treatment may suggest that only providing options is not enough to change behavior and that targeted education (public health) and information interventions may be necessary for healthy behavior change (Voth-Gaeddert et al., 2022d). These data help provide guidelines for building a larger framework as well as thinking about specific initiatives or programs to invest in as a stakeholder.

Water treatment program at scale

The UV water treatment pilot program provided an excellent model in which to evaluate component costs as well as costs at scale. The effectiveness of the UV treatment systems themselves suggested they are a viable option for producing potable water for households (i.e., Voth-Gaeddert and colleagues reported that 95.2% of post-treatment water samples had no detectable E. coli, N = 271). However, the cost-per-household given all of the components was $2,599, which if scaled to 10 or 50% of the total USVI population (on a per-household basis) would be $10.6 and $52.8 M, respectively. Fortunately, the pilot program was extensive in the services it offered allowing for a detailed breakdown of program costs and evaluation of where cost savings could be obtained by adapting the program structure to the needs of various populations across the islands.

The two most significant costs were the water treatment systems themselves and the water quality testing. At the wholesale price ($582), providing 10 or 50% of the islands’ households with a free UV system would cost $2.36 and $11.8 M, respectively (see Table 5). Targeting subsidies or vouchers to households whose primary barrier is the cost may help lower overall costs (as opposed to other barriers of knowledge of need or access mechanisms). In addition, total costs for hardware may be less based on the type of water treatment technology used. Water quality testing is important to ensure treatment system effectiveness, and significant cost savings could be obtained by integrating this program into an existing laboratory and reducing the frequency of sampling (e.g., four tests over 2 years vs. 12 tests over 1 year). This could reduce the testing cost closer to the per-sample cost of $9 per sample as capital lab costs and labor costs would be reduced at scale. Given a $30 per-sample cost and four free tests per household, total cost estimates would be $122k and $610k at 10 and 50% of households in USVI, respectively.

Table 5

Household water program costs at scale.

DescriptionCost at 10% of householdsCost at 50% of households
Water treatment system $2.36 M $11.8 M 
Water quality testinga $0.12 M $0.61 M 
System installations $1.63 M $8.13 M 
Annual replacement partsb $0.67 M $3.30 M 
Total program costsc $6.1 M $30.4 M 
DescriptionCost at 10% of householdsCost at 50% of households
Water treatment system $2.36 M $11.8 M 
Water quality testinga $0.12 M $0.61 M 
System installations $1.63 M $8.13 M 
Annual replacement partsb $0.67 M $3.30 M 
Total program costsc $6.1 M $30.4 M 

aMicrobiological water testing (no chemicals) assuming $30 per sample, four tests per household.

bTwo filters and one UV bulb (however, this would change if the water treatment system was different).

cIncludes 3 years of replacement parts and ignores administrative and management costs. Total USVI Households assumed to be 40,648.

Two other significant costs were for installations of the system ($400) and the annual replacement parts needed ($322 retail and $164 wholesale). While equitable plumber fees are important, cost sharing and program structuring may facilitate lower installation costs. In addition, bulk ordering may provide approaches to reduce or share costs for replacement parts. For installations, costs for 10 and 50% of households would be $1.63 and $8.13 M, respectively, while for annual replacement parts, it would be $0.67 and $3.3 M, respectively, at the wholesale price. While the above estimates come with a large margin of error, they represent the total cost of the whole territory-wide program, except for the cost of the annual replacement parts, program management labor, and an education campaign. Assuming 3 years of replacement parts are provided, the total program costs for reaching 10 and 50% of households would be $6.1 and $30.4 M, respectively. However, an information and education campaign is critical for driving demand for the services and user education for operation and maintenance. Furthermore, community health or case workers can be an effective option for enrollment, engagement, and support to users of the program but will increase management labor costs. While we do not have cost estimates for these two components, the available data can still be helpful in building out a larger multi-stakeholder framework. A coalition of stakeholders (i.e., households, government, NGOs, private, etc.) can use these values to evaluate how to allocate financial and management responsibilities across various stakeholder groups interested in improving water security throughout USVI.

While this study presents results important for decision-making, it is important to note the key limitations. First, the survey data only provided insights via two questions; therefore, as any program begins, a more substantial evaluation or feedback mechanism should be used to ensure the needs of the community are being addressed. Second, the water treatment pilot program utilized a UV system as the treatment technology; however, as we note, this estimated cost can be adjusted based on the desired water treatment system used in a territory-wide program. Finally, the water program costs at scale are only estimates based on the pilot program. We ignored several potential costs (due to a lack of data on costs for an education campaign and management) but also utilized higher cost estimates for certain components (e.g., water testing). The aim was to demonstrate that the order of magnitude of the costs for a scaled program is within the bounds of available finance. As any coalition explores a more detailed agenda for setting up a program, a more rigorous analysis should be conducted. Despite these limitations, the data here can provide useful information to local stakeholders and policymakers in establishing a territory-wide water storage and treatment program.

In this study, we aimed to leverage existing and new data to (1) identify critical components for a territory-wide program and (2) provide the relative costs of the core components necessary for the program. The survey data suggested the majority of households either did not know about the water treatment systems or did not know or think they needed one. In addition, among those that did know about the systems, the main barrier to acquiring one was the initial cost. In the evaluation of the UV system pilot program, the primary cost drivers were the actual water treatment system and water testing; however, strategic financing and targeted allocation can help provide equitable access at a manageable cost. The cost-per-household can be significantly reduced by optimizing and scaling the program. Finally, these data can be used to support efforts from stakeholders in understanding the level of financing necessary along with the types of resources and additional stakeholders necessary to be successful at establishing a territory-wide water storage and treatment program in USVI.

The authors are grateful for the staff at Love City Strong, Inc., as well as the volunteers and staff across various local and federal organizations that participated in the original cistern study. In addition, the authors would like to thank the guidance and support from Lieutenant Commander James Gooch.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

American Public Health Association (APHA)
(
2018
).
Standard Methods for the Examination of Water and Wastewater: Method 9223
.
American Public Health Association
,
New York, NY
.
Berkey Filters
(
2022
).
Berkey Light Water Filter – Shop Berkey Filters
.
[WWW Document]. Available at: https://www.berkeyfilters.com/products/berkey-light (Accessed March 3 2022)
.
Charlebois
B.
(
2021
).
Rain Flush Away: Evaluating Rainwater Catchment First Flush Volumes
.
University of Massachusetts, Amherst
.
Government of the United States Virgin Islands
(
2018
).
USVI Hurricane Recovery and Resilience Task Force: Water
.
Hamilton
K.
,
Reyneke
B.
,
Waso
M.
,
Clements
T.
,
Ndlovu
T.
,
Khan
W.
,
Digiovanni
K.
,
Montalto
F.
,
Haas
C. N.
&
Ahmed
W.
(
2019
).
A global review of the microbiological quality and potential health risks associated with roof-harvested rainwater tanks
.
npj Clean Water
2
(
7
), 1–18.
Imteaz
M.
&
Boulomytis
V.
(
2022
).
Improvement of rainwater harvesting analysis through an hourly timestep model in comparison with a daily timestep model
.
Water Resour. Manage.
36
,
2611
2622
.
https://doi.org/10.1007/S11269-022-03162-9/FIGURES/5
.
Imteaz
M. A.
&
Moniruzzaman
M.
(
2018
).
Spatial variability of reasonable government rebates for rainwater tank installations: a case study for Sydney
.
Resour. Conserv. Recycl.
133
,
112
119
.
https://doi.org/10.1016/J.RESCONREC.2018.02.010
.
Lindmark
M.
,
Cherukumilli
K.
,
Crider
Y.
,
Marcenac
P.
,
Lozier
M.
,
Voth-Gaeddert
L. E.
,
Lantagne
D. S.
,
Mihelcic
J. R.
,
Zhang
M.
,
Just
C.
&
Pickering
A. J.
(
2022
).
‘Passive chlorination for drinking water disinfection in resource-constrained settings: a critical review.’ Environmental Science & Technology 56 (13), 9164–9181. https://doi.org/10.1021/acs.est.1c08580
.
Rao
G.
,
Kahler
A.
,
Voth-Gaeddert
L. E.
,
Cranford
H.
,
Libby
S.
,
Galloway
R.
,
Molinari
N. -A.
,
Ellis
E. M.
,
Yoder
J.
,
Mattioli
M. C.
&
Ellis
B. R.
(
2022
).
Microbial characterization, factors contributing to contamination, and household use of US Virgin Island Cistern Water
.
Environ. Sci. Technol.
Environmental Science & Technology Water (accepted)
.
Seger
K. R.
,
Day
C. A.
,
Gaeddert
L.
,
Rao
G.
,
Kahler
A.
,
Ellis
E. M.
,
Ellis
B. R.
&
Byrd
B. D.
(
2022
).
A cross-sectional household survey in the US Virgin Islands (2019) reveals cisterns as challenging peridomestic Aedes aegypti habitats
.
J. Am. Mosq. Control Assoc.
38
,
113
117
.
https://doi.org/10.2987/21-7022
.
United States Census Bureau
(
2010
).
2010 Island Areas Censuses: U.S. Virgin Islands
.
US Department of Commerce, Washington, DC
.
Voth-Gaeddert
L. E.
&
Schranck
A.
(
2021
).
Inline chlorinator for potable water systems in low-resource settings
.
J. Environ. Eng.
147
(
7
),
2
5
.
Voth-Gaeddert
L. E.
,
Lemley
M.
,
Brathwaite
K.
,
Schranck
A.
&
Libbey
S.
(
2022a
). ‘
Design and Evaluation of a Household Chlorination System for Treating Cistern Water in the US Virgin Islands
.’
Journal of Environmental Engineering
148 (11), 1–4. https://doi.10.1061/(ASCE)EE.1943-7870.0002063
.
Voth-Gaeddert
L. E.
,
Lemley
M.
,
Willford
I.
,
Brathwaite
K.
,
Momberg
D.
,
Schranck
A.
&
Libbey
S.
(
2022b
). ‘
Evaluating a Large-Scale Batch Chlorination Method for Household Cistern Water Treatment in the US Virgin Islands
.’
Journal of Environmental Engineering 148 (10), 1–6. https://doi:10.1061/(ASCE)EE.1943-7870.0002059
.
Voth-Gaeddert
L. E.
,
Momberg
D.
,
Brathwaite
K.
,
Schranck
A.
,
Lemley
M.
&
Libbey
S.
(
2022c
).
‘Household UV Treatment of Cistern Water as a Viable Option for Generating Potable Water in the US Virgin Islands.’ in review
.
Voth-Gaeddert
L. E.
,
Momberg
D.
,
Nkambule
B.
,
Rohloff
P.
,
Said-Mohamed
R.
&
van Heerden
A.
(
2022d
).
‘A Framework for Empowering Marginalized Communities via Information Interventions in Global Health: Examples from WASH and Malnutrition.’ in review
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data