Grantee Research Project Results
2019 Progress Report: Development of a Community-Based Lead Risk and Mitigation Model
EPA Grant Number: CR839376Title: Development of a Community-Based Lead Risk and Mitigation Model
Investigators: Cuppett, Jonathan
Institution: Water Research Foundation
Current Institution: Water Research Foundation , Cornwell Research Group , University of Florida
EPA Project Officer: Hahn, Intaek
Project Period: May 1, 2018 through April 30, 2021 (Extended to April 30, 2024)
Project Period Covered by this Report: May 1, 2019 through April 30,2020
Project Amount: $1,981,000
RFA: National Priorities: Transdisciplinary Research into Detecting and Controlling Lead in Drinking Water (2017) RFA Text | Recipients Lists
Research Category: Drinking Water , Endocrine Disruptors , Water
Objective:
The objective of this research is to further the science of risk assessment and mitigation for lead (Pb) exposure, and to translate this into practices that limit or even eliminate Pb exposure for drinking water consumers, especially children and pregnant women. This will be accomplished through these specific tasks: (1) generation of a quantitative lead risk based computational model (QLRM) that is built on a comprehensive and enhanced national dataset, (2) identification of opportunities to mitigate Pb exposure in drinking water, and (3) development of a communication framework to educate stakeholders on risk mitigation opportunities.
Progress Summary:
The development of the Lead Risk and Mitigation Model proceeded during this reporting period. The model requires a number of modules, and central to the model is the module to convert water lead levels (WLL) to blood lead levels (BLL) and then to IQ impacts. During this year that module was developed, reviewed by the research team and a spreadsheet calculation was developed for the conversions.
Similarly, work proceeded on the engineering side of the model. That is, analysis of datasets to determine factors which affect the WLL in water consumed. Those modules include development of the solubility models for lead release from piping material, establishment of the relationship between LCR compliance data and published BLLs of children and lead source identification through lead isotopes. The latter task requires the collection of blood as well as environmental samples for lead isotope analysis; during this period, the team worked to establish partner protocols and sampling requirements. The research team worked during this time to develop a database of lead isotopes and isotope ratios from various waters, plumbing materials, and pipe scales from a number of water systems. This database will help inform the model input regarding lead sources in drinking water.
Concurrently, the research team is developing the web-based application that will allow general utilization of the model. The user interface will primarily be used for visual design and layout and serve as a way for the user to interact with the calculations programmed into the microservices and related database.
The Quality Assurance Project Plan (QAPP) was approved by EPA on May 13, 2019. Since that date, the project team initiated work on a number of tasks related to the development of the risk based computational model. A meeting was held with the entire research team at the University of Florida on September 24th, 2019. The meeting served to better define the model structure, the inputs/outputs of the model and the role of each researcher. Additional details, by project task, are described below. At this time, the original project goals are being pursued. As described below, COVID-19 related restrictions are beginning to impact multiple tasks. The severity and duration of the impact is unknown at this time. We will communicate with our EPA project officer if these restrictions begin to impact our project goals and outputs/outcomes. We continue to work with our project partners on IRB and Human Subjects documentation. We anticipate submitting IRB documents to EPA for approval during the next reporting period
Task 1. Quantitative Lead Risk Model (QLRM)
The model development requires several modules to be created. Central to the model is the module to convert water lead levels (WLL) to blood lead levels (BLL) and then to IQ impacts. During this year that module was developed and a spreadsheet calculation was developed for the conversions. That module is described in Task 1a. Another module is the WLL input. One basic input will be the measured WLL generally reported as the 1st liter draw as required under the lead and copper rule (LCR). One of the goals of the input module will be to convert the input WLL to a time weighted exposure. That work is just starting. A third module is the mitigation routine that accepts an action and estimates the change in WLL. That development is ongoing.
Task 1a. Development of Health Outcomes Module
Dr. Crawford Brown continues to lead the work on the model, particularly on the health outcomes. The effort in this year was in developing the structure of the model, including inputs required, outputs generated and sensitivity analysis capability. A prototype structure was developed, focused for now on IQ impacts. Other health effects may be added at a later date as desired, although these will not fundamentally change the structure the model (parameter values will however change).
Task 1b. Development of Engineering Side of the Model (Mitigation impacts on WLL)
The overall objective of this task is to improve prediction of WLL through the model in the mitigation step. It is especially useful in the mitigation module for estimating changes in lead levels based on mitigation strategies employed. The approach is to collect and utilize existing data, supplemented with additional data collection to build relationships among the variables that when coupled with probabilistic analysis allows for better WLL prediction. The following subtasks are designed to help develop these inputs.
Task 1 b 1. Forensic Lead Isotope Analysis
Contact and follow up calls have been made with a number of potential partners to obtain subjects with elevated blood lead levels along with paired environmental samples. The Pueblo Department of Public Health and Environment has agreed to partner in our study and Butte, MT is still considering partnering. DC Water declined to participate. In February, 2020, we were informed that the Galesburg, IL study did not identify a sufficient number of subjects with blood levels higher than 5 µg/dL and so that research team determined that it could not participate with us on this study. Pueblo can likely provide all of the subjects needed for the study, although if Butte or another location agree to partner, subjects will be obtained from all partners.
We are working on IRB approvals and protocols for Pueblo and our other partners. Each partner is working with us individually, and when we complete each process, we will submit all information to EPA for approval before any samples are actually collected. A site visit and presentation with the “Colorado Lead Coalition” was scheduled for March 2020. However, both the Coalition meeting and the visit were postponed due to COVID-19 concerns.
Task 1 b 2. Laboratory studies on Pb isotopes
During this period, we conducted Pb isotope analyses on water, drinking water pipes, and pipe scales samples. Preliminary results obtained from these lab efforts were presented on a poster at the American Geophysical Union fall meeting in December 2019.
Task 1 b 3. Literature review
A review of the literature addressing the link between lead (Pb) levels in environmental samples and Pb in human blood is ongoing at UF. This review is intended for publication; it seeks to summarize published data on the difference between Pb from different environmental sources (e.g. soil, water, air, paint, food, etc.) and children’s BLLs. With the ongoing confinement due to COVID-19, and current UF’s guidelines requesting that all work be done from home, laboratory studies have been halted. This pause frees time to finalize the manuscript for publication. We plan to complete this effort by the end of May 2020. We also completed the “high level” comparison of CDC reported blood lead data compared to utility reported water lead data to look for any relationships. With the dataset available, we did not find a correlation.
Task 1 b 4. Laboratory work to verify the utility of solubility models for Pb scale dissolution
UF has initiated laboratory studies on the predictions of the dissolution of Pb from scale materials with known solid phase mineral composition as a function of water chemistry. The goal of this exercise is to verify the ability of geochemical equilibrium models such as MINTEQ to accurately predict the dissolution of Pb as a function of changing water chemistry. Because of COVID-19 concerns, UF has requested that all work be done from home. Therefore laboratory studies on the predictions of the dissolution of Pb from scale materials as a function of water chemistry was stopped as of April 2, 2020.
Task 1 b 5. Geochemical modeling to predict Pb solubility as a function of key water chemistry parameters
This task seeks to predict the dissolution of Pb from drinking water pipes based on knowledge of water chemistry and prevalent Pb-mineral phases present in scales. In this research component, we are using XRD data on prevalent Pb-minerals present in scales of drinking water pipes, and the chemical composition of water samples collected for the pipes’ locations to predict Pb release to water as a function of changing key water quality parameters and the identified Pb-mineral phases controlling the dissolution of Pb.
Task 1 b 6. Laboratory studies of the dissolution of Pb from drinking water pipes’ scales
For this first set of laboratory experiments, lead pipes and drinking water samples collected from the same location are used. Water of known chemical composition (used in modeling studies) is equilibrated with scales within a 1-ft piece of lead pipe, under static conditions. In addition to testing water with different chemical compositions, the effect of equilibration time will be investigated. For both the modeling and experimental components, key water chemistry parameters to be tested include pH, DIC, hardness (competitive cations), phosphate, and DOC (e.g. ligands with electron donor atoms). Generated samples will be analyzed by ICP-MS and trends of Pb dissolution compared to those predicted using modeling tools. These experiments are currently halted due to the COVID-19 imposed confinement.
Task 1 b 7. Home level research on lead release
CRG is planning research in an abandoned home in Flint, MI to better understand which parameters influence the release of lead in a home, thereby contributing to the WLL. CRG reconfirmed that the Flint Land Bank will work with us on this project in a meeting held December 17, 2019. The Land Bank currently owns the title to nearly 3000 homes within the Flint City limits. CRG identified these properties and have asked the City of Flint personnel to help choose a property that has a lead service line, since these homes were not part of the project to replace lead service lines. Once a property is selected, CRG will begin a sampling regime including first draw and sequential sampling (small and large volume), as well as develop a sampling protocol that takes advantage of the opportunity to conduct research in this home that will help identify the impact of premise planning on lead release.
Task 1c. Web Application Development
Concurrent with the development of the Lead Risk and Mitigation Models, the development team will create a web-based application to provide a communication framework that exposes the functionality created by the Research Team. A web meeting was held in March, 2020 with the research team and the web development team to assure that the expectations of the research team can be accomplished through the work of the web development team. The core functionality of the web application will include:
(1) Creation of a security / user management framework
(2) Wizard style input method, with display of resulting model outputs
(3) Ability to save work from each model as inputs into related lead models
(4) Simulated outputs of health outcomes and associated costs
1 c 1. Proposed Technical Architecture
The Excel models being developed by the research team will be converted into an online web application that can be run from anywhere. The data fields in the Excel model will be recreated in a database, SQL Server or SQL Lite, and then exposed to a user interface through the .NET programming framework. The web development team plans to leverage .NET Microservices to communicate between the database and user interface. The calculations for the model will be done either at the Microservice level or through JavaScript on the front-end of the user interface. Default reference numbers found in the model will be stored in configuration files, so they are easy to change, if needed.
1 c 2. Microservices Architecture
A microservices architecture is an approach to development that uses a set of small encapsulated services, with core business functionality primarily oriented to the back-end. Each of the Excel models will be converted into their own microservice and contain calculations / functions currently found in the Excel models. Each service communicates through an end-to-end domain or business capability using HTTP/HTTPs. Each microservice shares a related data model and allows for greater flexibility and scale. The framework uses traditional HTTP as the communication protocol between the client app and microservices and supports asynchronous (real-time) data updates.
1 c 3. User Interface
The majority of the calculations and model functionality will be contained in the microservice layer. The user interface will primarily be used for visual design and layout and serve as a way for the user to interact with the calculations programmed into the microservices and related database. The user interface / front-end environment is programmed in Microsoft Razor (template syntax with embedded HTML), with additional front-end languages and/or frameworks, such as JavaScript, JQuery, React or Angular used as necessary.
1 c 4. Database
The model inputs and related outputs are stored in an online database. The database will be developed in SQL Server or SQL Lite. The team is initially targeting Azure SQL for the database, but the functionality will be portable to any hosting platform capable of hosting a .NET based application.
1 c 5. Hosting
The entire platform will be portable and able to be hosted in multiple environments including cloud based hosting on Microsoft Azure, traditional web hosting, or on-premises hosting on a Windows or Linux environment.
1 c 6. Technical Approach of Web application development
The project is structured around four primary tasks: designing the user interface for the web application, developing the IQ model as a proof-of-concept (POC), stakeholder feedback on the IQ model, and incorporating additional models into the IQ model.
Task 2. Lead Reduction Strategies
Work on this task will begin in the next report period.
Task 3. Education and Outreach
We continue to communicate with our communication partner American Water Works Association (AWWA). As actionable information is developed we will work with AWWA to conduct education and outreach to appropriate stakeholders.
Future Activities:
During the next project period, work will focus on the inputs of the model. It is assumed, as indicated by UF, that all laboratory work at UF will resume by June. Concerns over the Covid virus has resulted in the UF laboratories closing, but the UF researchers are working remotely, and all CRG laboratory work is continuing. All of the research team continues to work during this time.
Specific tasks include:
1. Continued work on the web application of the model.
2. Completion of the IRB process with the UF and submittal to EPA, leading to the collection of samples of blood and environmental samples for the lead isotope work.
3. Completion of the acquisition of access to the Flint home and conducting sampling at the home.
4. Data assembly and organization into a compatible form for use by the model, including the sequential sampling lead data base, and national lead occurrence databases.
5. Determine the reduction capability of lead removal devices.
6. Conduct a meeting focused on the model and its web application during the summer of 2020.
7. Hold a workshop in 2020 to gather feedback on the initial design of the model.
Journal Articles:
No journal articles submitted with this report: View all 6 publications for this projectSupplemental Keywords:
Lead, modeling, drinking water, corrosion, service lines, human health effects, human exposure, risk mitigation, metals, lead sources, infrastructure, communication, blood lead level, water lead level, lead isotope analysisRelevant Websites:
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- Final Report
- 2022 Progress Report
- 2021 Progress Report
- 2020 Progress Report
- 2018 Progress Report
- Original Abstract
3 journal articles for this project