How a UW tool helps Milwaukee decide which lead pipes to replace first

Reading Time: 6 minutes

Milwaukee faces a big challenge: replacing tens of thousands of lead water pipes that pose a potential health risk to residents — all by a 2037 deadline imposed by the U.S. Environmental Protection Agency. 

At the end of 2025, Milwaukee Water Works still had over 55,000 residential homes or properties with lead water lines on the utility-owned side, customer-owned side or both. The utility had yet to identify the makeup of pipelines serving another 27,000 properties. 

But Milwaukee can’t replace all the pipelines at once — and some neighborhoods face greater risks from lead exposure than others.

So which areas should officials prioritize? 

Part of the answer lies in a surprising place: a tool University of Wisconsin-Madison researchers originally developed to study the connection between neighborhood disadvantage and human health.

Lead does not impact neighborhoods equally

div, body, .container, .entry-content, .wp-block-group__inner-container, .wp-block-group, .wp-block-column {
overflow: visible !important;
overflow-x: visible !important;
overflow-y: visible !important;
}

const stepOneImage = “https://wisconsinwatch.org/wp-content/uploads/2026/10/Step1-oldmap_Desktop-scaled.png”;
const stepTwoImage = “https://wisconsinwatch.org/wp-content/uploads/2026/10/Step2-inventory_Desktop-scaled.png”;
const stepOneImageMobile = “https://wisconsinwatch.org/wp-content/uploads/2026/10/Step1-oldmap_Mobile-scaled.png”;
const stepTwoImageMobile = “https://wisconsinwatch.org/wp-content/uploads/2026/10/Step2-inventory_Mobile-scaled.png”;

Milwaukee’s pipeline problem dates back to 1872, when the city mandated the use of lead in service lines. Although the city abolished the mandate in 1962, older lead pipelines remained in many older residential buildings.

Downtown Milwaukee includes a scattering of homes with non-lead service lines — many due to having already been replaced by the city.

.step-change-map {
width: 100%;
height: auto;
display: block;
}

.step-change-media {
position: sticky;
top: 0;
z-index: 0;
}

.step-change-media img, .step-change-media figcaption {
margin-left: auto;
margin-right: auto;
}

.step-change-map–one,
.step-change-map–two {
transition: opacity 260ms linear;
}

.step-change-map–one {
opacity: 1;
}

.step-change-map–two {
position: absolute;
inset: 0;
opacity: 0;
pointer-events: none;
}

.sticky-block {
position: relative;
–sticky-header-offset: 205px;
margin-bottom: 32px;
}

.step-change-footnote {
display: grid;
margin-top: 8px;
}

#step-change-footnote-one,
#step-change-footnote-two {
grid-area: 1 / 1;
display: block;
transition: opacity 260ms linear;
}

#step-change-footnote-one {
opacity: 1;
}

#step-change-footnote-two {
opacity: 0;
}

#step-change-text {
position: relative;
z-index: 1;
margin-top: 160px;
width: 90%;
margin-left: auto;
margin-right: auto;
opacity: 0;
transform: translateY(18px);
will-change: opacity, transform;
}

.sticky-spacer {
height: calc(100vh + var(–sticky-header-offset));
}

#step-change-text p {
padding-left: 1rem;
padding-right: 1rem;
}

.liquidglass{
background: rgba(255, 255, 255, 0.5);
border: 1px solid rgba(255, 255, 255, 0.2);
border-radius: 25px;
-webkit-backdrop-filter: blur(16px) saturate(1.6);
backdrop-filter: blur(16px) saturate(1.6);
box-shadow:
inset 0 1px 0 rgba(255, 255, 255, 0.55),
0 10px 28px rgba(0, 0, 0, 0.16);
}

@supports not ((-webkit-backdrop-filter: blur(1px)) or (backdrop-filter: blur(1px))) {
.liquidglass {
background: rgba(255, 255, 255, 0.85);
}
}

@media (max-width: 600px) {
.section-wrapper {
padding-top: clamp(20px, 5vh, 60px);
}
}

(() => {
const stickyBlock = document.querySelector(“.sticky-block”);
const stepMedia = document.querySelector(“.step-change-media”);
const stepText = document.getElementById(“step-change-text”);
const stepImageOne = document.getElementById(“step-change-image-one”);
const stepImageTwo = document.getElementById(“step-change-image-two”);
const stepFootnote = stepMedia ? stepMedia.querySelector(“.step-change-footnote”) : null;
const stepFootnoteOne = document.getElementById(“step-change-footnote-one”);
const stepFootnoteTwo = document.getElementById(“step-change-footnote-two”);

if (!stickyBlock || !stepMedia || !stepText || !stepImageOne || !stepImageTwo || !stepFootnote || !stepFootnoteOne || !stepFootnoteTwo) {
return;
}

const clamp = (value, min, max) => Math.max(min, Math.min(max, value));
const mobileThresholdRatio = 4.5 / 8;
const mobileHeaderOffset = 145;
let isTicking = false;
let activeStepOneSrc = “”;
let activeStepTwoSrc = “”;

const getHeaderOffset = () => {
const headerOffsetValue = getComputedStyle(stickyBlock).getPropertyValue(“–sticky-header-offset”).trim();
const parsedHeaderOffset = Number.parseFloat(headerOffsetValue);
return Number.isFinite(parsedHeaderOffset) ? parsedHeaderOffset : 155;
};

const getImageSources = () => {
const viewportRatio = window.innerWidth / window.innerHeight;
const hasMobileSources =
typeof stepOneImageMobile === “string” &&
stepOneImageMobile.length > 0 &&
typeof stepTwoImageMobile === “string” &&
stepTwoImageMobile.length > 0;
const useMobile = hasMobileSources && viewportRatio {
const { useMobile } = getImageSources();
const headerOffset = useMobile ? mobileHeaderOffset : getHeaderOffset();
const viewportHeight = window.innerHeight;
const availableHeight = Math.max(0, viewportHeight – headerOffset);
const mediaHeight = stepMedia.getBoundingClientRect().height;
const footnoteRect = stepFootnote.getBoundingClientRect();
const footnoteStyles = window.getComputedStyle(stepFootnote);
const footnoteMarginTop = Number.parseFloat(footnoteStyles.marginTop) || 0;
const footnoteMarginBottom = Number.parseFloat(footnoteStyles.marginBottom) || 0;
const footnoteTotalHeight = footnoteRect.height + footnoteMarginTop + footnoteMarginBottom;
const maxImageHeight = Math.max(0, availableHeight – footnoteTotalHeight);

if (!useMobile && availableHeight > 0 && mediaHeight > availableHeight) {
stepMedia.style.maxHeight = `${availableHeight}px`;
stepImageOne.style.height = `${maxImageHeight}px`;
stepImageTwo.style.height = `${maxImageHeight}px`;
stepImageOne.style.width = “auto”;
stepImageTwo.style.width = “auto”;
stepImageOne.style.maxWidth = “100%”;
stepImageTwo.style.maxWidth = “100%”;
} else {
stepMedia.style.maxHeight = “”;
stepImageOne.style.height = “auto”;
stepImageTwo.style.height = “auto”;
stepImageOne.style.width = “100%”;
stepImageTwo.style.width = “100%”;
stepImageOne.style.maxWidth = “”;
stepImageTwo.style.maxWidth = “”;
}

const adjustedMediaHeight = stepMedia.getBoundingClientRect().height;
const centeredTop = headerOffset + Math.max(0, (availableHeight – adjustedMediaHeight) / 2);
stepMedia.style.top = `${useMobile ? headerOffset : centeredTop}px`;
};

const applyState = () => {
const rect = stickyBlock.getBoundingClientRect();
const totalTravel = rect.height + window.innerHeight;
const progress = clamp((window.innerHeight – rect.top) / totalTravel, 0, 1);
const { stepOneSrc, stepTwoSrc } = getImageSources();
let sourceChanged = false;

if (stepOneSrc !== activeStepOneSrc) {
stepImageOne.src = stepOneSrc;
activeStepOneSrc = stepOneSrc;
sourceChanged = true;
}

if (stepTwoSrc !== activeStepTwoSrc) {
stepImageTwo.src = stepTwoSrc;
activeStepTwoSrc = stepTwoSrc;
sourceChanged = true;
}

if (sourceChanged) {
updateStickyTop();
}

let opacity = 0;
if (progress >= 0.12 && progress <= 0.88) {
if (progress 0.65) {
opacity = (0.88 – progress) / (0.88 – 0.65);
} else {
opacity = 1;
}
}
opacity = clamp(opacity, 0, 1);

stepText.style.opacity = String(opacity);
stepText.style.transform = `translateY(${(1 – opacity) * 18}px)`;

const crossfadeStart = 0.35;
const crossfadeEnd = 0.65;
const stepTwoOpacity = clamp((progress – crossfadeStart) / (crossfadeEnd – crossfadeStart), 0, 1);
const stepOneOpacity = 1 – stepTwoOpacity;
stepImageOne.style.opacity = String(stepOneOpacity);
stepImageTwo.style.opacity = String(stepTwoOpacity);
stepFootnoteOne.style.opacity = String(stepOneOpacity);
stepFootnoteTwo.style.opacity = String(stepTwoOpacity);
stepFootnoteOne.style.pointerEvents = stepOneOpacity > 0.5 ? “auto” : “none”;
stepFootnoteTwo.style.pointerEvents = stepTwoOpacity > 0.5 ? “auto” : “none”;
};

const onScrollOrResize = () => {
if (isTicking) {
return;
}

isTicking = true;
requestAnimationFrame(() => {
applyState();
isTicking = false;
});
};

const onResize = () => {
updateStickyTop();
onScrollOrResize();
};

const applyLayoutNow = () => {
updateStickyTop();
applyState();
};

window.addEventListener(“scroll”, onScrollOrResize, { passive: true });
window.addEventListener(“resize”, onResize);
stepImageOne.addEventListener(“load”, updateStickyTop);
stepImageTwo.addEventListener(“load”, updateStickyTop);

if (stepImageOne.complete && stepImageTwo.complete) {
requestAnimationFrame(applyLayoutNow);
} else {
window.addEventListener(“load”, () => {
requestAnimationFrame(applyLayoutNow);
}, { once: true });
applyLayoutNow();
}
})();

Milwaukee has replaced nearly 15,000 pipelines since beginning its broader efforts in 2017, said Milwaukee Water Works Superintendent Patrick Pauly. When federal funding became available under the Bipartisan Infrastructure Law, the utility expanded its effort, increasing the pace from about 1,000 lines a year to 5,000. 

The expansion included a priority program launched in 2024, which focuses on equity and offers free replacements to eligible residents. 

But scaling up the effort required more than funds and construction crews. The city also needed a systematic way to decide where to work — one that recognizes disproportionate risks across the city.

“We needed to expand the lead service line replacement program, and we needed to develop a, for lack of a better word, defensible method to expand that program,” Pauly said.

What Milwaukee’s priority system looks like

The resulting system assigned each property a score based on three factors: 

• 70% comes from the Area Deprivation Index (ADI), a measure of relative socioeconomic disadvantage developed by UW-Madison researchers. 
• 25% comes from the percentage of local children who tested positive for lead poisoning. 
• 5% comes from lead service line density in the area.

Together, these measures help the utility determine where replacement will yield the highest benefits. 

The city updates each property’s score every other year as new information becomes available, Pauly said. 

The scoring weights heavily toward neighborhood disadvantage as broader challenges residents face can compound risks posed by lead exposure. 

Nationwide, communities burdened by economic disadvantages and racial inequalities can face greater risks of lead exposure, according to a Montclair State University study of New Jersey. Such neighborhoods often contain older housing and sit near former industrial sites.

The ADI relies on attempts to quantify disadvantages in Milwaukee, and its origins have little to do with water infrastructure.

A tool developed to understand health disparities

The ADI measures neighborhood socioeconomic disadvantage or adverse social exposure using 17 U.S. Census American Community Survey measures. They include factors related to income, education, housing, and employment.

The index assigns scores to census block groups — small geographic areas that generally contain several hundred to a few thousand people.

Block groups receive a national percentile ranking from 1 to 100 or a statewide decile ranking from 1 to 10. Higher rankings indicate greater socioeconomic disadvantage relative to other areas.

#g-ADIMap-box {
container-type:inline-size;
container-name:g-ADIMap-box;
}
#g-ADIMap-box,
#g-ADIMap-box .g-artboard {
margin:0 auto;
}
#g-ADIMap-box p {
margin:0;
}
#g-ADIMap-box .g-aiAbs {
position:absolute;
}
#g-ADIMap-box .g-aiImg {
position:absolute;
top:0;
display:block;
width:100% !important;
}
#g-ADIMap-box .g-aiSymbol {
position:absolute;
box-sizing:border-box;
}
#g-ADIMap-box .g-aiPointText p {
white-space:nowrap;
}
#g-ADIMap-Desktop {
position:relative;
overflow:hidden;
}
#g-ADIMap-Desktop {
display:none;
}
@container g-ADIMap-box (width >= 780px) {
#g-ADIMap-Desktop {
display:block;
}
}
#g-ADIMap-Desktop p {
font-family:’Heebo’, sans-serif;
font-weight:regular;
line-height:22px;
opacity:1;
letter-spacing:0em;
font-size:18px;
text-align:left;
color:rgb(0,0,0);
text-transform:none;
padding-bottom:0;
padding-top:0;
mix-blend-mode:normal;
font-style:normal;
height:auto;
position:static;
}
#g-ADIMap-Desktop .g-pstyle0 {
font-weight:700;
line-height:24px;
font-size:20px;
}
#g-ADIMap-Desktop .g-pstyle1 {
font-weight:400;
}
#g-ADIMap-Desktop .g-pstyle2 {
font-weight:700;
line-height:17px;
font-size:14px;
text-align:center;
}
#g-ADIMap-Desktop .g-pstyle3 {
font-weight:400;
line-height:14px;
font-size:12px;
}
#g-ADIMap-Desktop .g-pstyle4 {
font-weight:400;
text-align:right;
}
#g-ADIMap-Desktop .g-cstyle0 {
font-weight:700;
}
#g-ADIMap-Mobile {
position:relative;
overflow:hidden;
}
@container g-ADIMap-box (width >= 780px) {
#g-ADIMap-Mobile {
display:none;
}
}
#g-ADIMap-Mobile p {
font-family:’Heebo’, sans-serif;
font-weight:regular;
line-height:19px;
opacity:1;
letter-spacing:0em;
font-size:16px;
text-align:left;
color:rgb(0,0,0);
text-transform:none;
padding-bottom:0;
padding-top:0;
mix-blend-mode:normal;
font-style:normal;
height:auto;
position:static;
}
#g-ADIMap-Mobile .g-pstyle0 {
font-weight:700;
line-height:22px;
font-size:18px;
}
#g-ADIMap-Mobile .g-pstyle1 {
font-weight:400;
}
#g-ADIMap-Mobile .g-pstyle2 {
font-weight:700;
text-align:center;
}
#g-ADIMap-Mobile .g-pstyle3 {
font-weight:400;
text-align:right;
}
#g-ADIMap-Mobile .g-cstyle0 {
font-weight:700;
}

Researchers have long recognized that people’s health depends on far more than the medical care they receive. But measuring the influence of social and economic conditions on health isn’t easy.

Dr. Amy Kind has spent much of her career trying to figure that out.

Kind, a professor of geriatrics and gerontology at UW-Madison and founding director of its Center for Health Disparities Research, grew up in Niagara, Wisconsin, before training as a physician. 

During her residency in Boston, she saw that the medicine she prescribed could do only so much when patients returned to homes without healthy food, stable housing or access to other basic resources. Those experiences helped shape her interest in finding ways to measure such disadvantages. 

Researchers have used indexes like ADI since the 1980s. In 2014, Kind led a group of University of Wisconsin School of Medicine and Public Health researchers in updating ADI using modern census data, refined it to the census block level and validated its usefulness across several scientific studies. 

“(Like any marker), it may not be the thing that ultimately drives the risk,” Kind said. “But it has been an extremely strong marker that something going on in that area is increasing risk, whatever it might be.”

An unexpected use for health research

“When we first developed the metric, we thought that perhaps it would be used primarily by health systems, primarily in the research community,” Kind said. 

Instead, Kind and her colleagues heard from many community groups and local governments that were interested in using index data to guide their public health work.

Since 2014, the University of Wisconsin has tracked over 3,500 academic papers citing the index and more than 400 examples of its use across different fields. That includes a recent Kind study linking environmental factors to the likelihood of developing Alzheimer’s disease. Meanwhile, the index helped officials in Puerto Rico decide where to send nurses to combat dengue fever and the mosquitoes that carry it.

That wide interest prompted Kind to develop the Neighborhood Atlas in 2018, making the data more accessible. 

Kind said she enjoys seeing the index be used in sometimes unexpected ways to address local challenges. That includes Milwaukee’s effort to protect children from lead exposure, where officials are using it as one component of their model. 

Thousands of pipelines replaced, thousands more to go 

Milwaukee Water Works has replaced lead pipes at more than 6,000 properties under its priority program, which launched less than three years ago.

#g-Taskforce-box {
container-type:inline-size;
container-name:g-Taskforce-box;
}
#g-Taskforce-box,
#g-Taskforce-box .g-artboard {
margin:0 auto;
}
#g-Taskforce-box p {
margin:0;
}
#g-Taskforce-box .g-aiAbs {
position:absolute;
}
#g-Taskforce-box .g-aiImg {
position:absolute;
top:0;
display:block;
width:100% !important;
}
#g-Taskforce-box .g-aiSymbol {
position:absolute;
box-sizing:border-box;
}
#g-Taskforce-box .g-aiPointText p {
white-space:nowrap;
}
#g-Taskforce-Desktop {
position:relative;
overflow:hidden;
}
#g-Taskforce-Desktop {
display:none;
}
@container g-Taskforce-box (width >= 780px) {
#g-Taskforce-Desktop {
display:block;
}
}
#g-Taskforce-Desktop p {
font-family:’Heebo’, sans-serif;
font-weight:regular;
line-height:24px;
opacity:1;
letter-spacing:0em;
font-size:20px;
text-align:left;
color:rgb(0,0,0);
text-transform:none;
padding-bottom:0;
padding-top:0;
mix-blend-mode:normal;
font-style:normal;
height:auto;
position:static;
}
#g-Taskforce-Desktop .g-pstyle0 {
font-weight:700;
}
#g-Taskforce-Desktop .g-pstyle1 {
font-weight:700;
line-height:22px;
font-size:18px;
text-align:center;
}
#g-Taskforce-Desktop .g-pstyle2 {
font-weight:400;
line-height:22px;
font-size:18px;
}
#g-Taskforce-Desktop .g-pstyle3 {
font-weight:400;
line-height:22px;
font-size:18px;
text-align:right;
}
#g-Taskforce-Mobile {
position:relative;
overflow:hidden;
}
@container g-Taskforce-box (width >= 780px) {
#g-Taskforce-Mobile {
display:none;
}
}
#g-Taskforce-Mobile p {
font-family:’Heebo’, sans-serif;
font-weight:regular;
line-height:22px;
opacity:1;
letter-spacing:0em;
font-size:18px;
text-align:left;
color:rgb(0,0,0);
text-transform:none;
padding-bottom:0;
padding-top:0;
mix-blend-mode:normal;
font-style:normal;
height:auto;
position:static;
}
#g-Taskforce-Mobile .g-pstyle0 {
font-weight:700;
}
#g-Taskforce-Mobile .g-pstyle1 {
font-weight:700;
line-height:19px;
font-size:16px;
text-align:center;
}
#g-Taskforce-Mobile .g-pstyle2 {
font-weight:400;
line-height:19px;
font-size:16px;
}
#g-Taskforce-Mobile .g-pstyle3 {
font-weight:400;
line-height:19px;
font-size:16px;
text-align:center;
}
#g-Taskforce-Mobile .g-pstyle4 {
font-weight:400;
line-height:19px;
font-size:16px;
text-align:right;
}

The work has been largely concentrated on the city’s North Side but has also expanded into near South Side neighborhoods. 

Pauly expects the utility to reach a milestone next year of 20,000 lead lines replaced.

But officials will soon face a difficult question: how to pay for replacements after  Bipartisan Infrastructure Law funding runs out in 2028.  

One possibility is returning to a funding system that required homeowners to pay part of the cost of replacing the privately owned portion of their service lines. 

“I don’t believe the city has the financial capacity to take on fully funding the private side replacement costs,” Pauly said.

That could present another challenge for a program designed to prioritize neighborhoods where residents already face financial disadvantages.

For now, residents living in homes with lead service lines that are not yet scheduled for replacement can take steps to better understand their own risk. 

Julie LaRose, lead and nutrition program coordinator at the MacCanon Brown Homeless Sanctuary, runs one of the many programs in the city where residents can submit samples of their tap water for free lead testing.

LaRose recognizes residents’ concerns about their tap water. But she also emphasizes that Milwaukee water is highly regulated, and she has not seen any samples tested through her program exceed one part per billion (ppb) of lead. 

That’s far below the FDA limit for lead in bottled water (5 ppb) and the EPA’s threshold for utility action (15 ppb). 

“If you’re nervous about this, let’s test it,” LaRose said.

How a UW tool helps Milwaukee decide which lead pipes to replace first is a post from Wisconsin Watch, a non-profit investigative news site covering Wisconsin since 2009. Please consider making a contribution to support our journalism.