Data Center Lawsuit Settlements and Verdicts: Every Dollar Figure on Record, and the Analogous Cases That Predict the Next Ones

Two payments exist, both from the same case, and only one from a data center company.

Defendant	Amount	Date	What it resolves	Caveat
Amazon Data Services, Inc.	$20.5 million	Motion for preliminary approval filed March 31, 2026 (D. Or.)	Class claims that wastewater from Amazon's 13 eastern Oregon data centers contributed to nitrate contamination of groundwater in the Lower Umatilla Basin	Amazon denies contributing to the nitrate problem. Funds go to well and water infrastructure, not individual checks. Subject to court approval.
Madison Ranches	$6.2 million	Announced August 12, 2026	The same class claims, against a large farm in the same basin	An agricultural defendant, not a data center. Does not admit wrongdoing. Money funds filtration, deeper wells, or city water hookups for well-dependent residents.

The Oregon settlement is the benchmark for every data center water claim, and its structure matters as much as its size. It is a class action covering a defined groundwater management area, the harm is regional rather than house by house, and the remedy is fixing the water. The case continues against the remaining defendants, including Lamb Weston, the Port of Morrow, Threemile Canyon Farms, Portland General Electric, and Columbia River Processing, with a trial scheduled for 2027.

That is the whole record. The Microsoft Fairwater noise class, the Southaven turbine class, the Dowagiac class, the Louisa County well suit, and the Newton County suit against Meta are all pending. The Saline Township consent judgment included a community benefits package for the township, but that was the price of a project going forward, not compensation to injured neighbors. Our by the numbers report counts every case.

Data center neighbor claims are private nuisance, negligence, and trespass claims, the same theories that have been tried against hog operations, landfills, smelters, chemical plants, and energy construction sites for decades. Those cases are the real reference points, and the figures below come from court opinions, filings, and the public results pages of the firms that handled them. Read the caveat column before the amount.

Case	Industry and harm	Result	Caveat
H&L Farms v. Silicon Ranch Corp., M.D. Ga., Columbus Division (No. 4:21-cv-134)	Sediment and erosion from an upstream solar construction site damaged a 21 acre lake in Stewart County, Georgia; nuisance and negligence	Jury verdict, April 28, 2023: $135.5 million, made up of $4.5 million compensatory to each of two homeowners plus $125 million punitive against Silicon Ranch and two construction contractors	Two plaintiffs, one destroyed lake, and 92 percent of the figure is punitive. On October 23, 2023 the trial judge upheld liability and the specific intent findings but remitted the total to roughly $5.1 million, the plaintiffs rejected the remittitur, and the case settled for an undisclosed amount in January 2025 with no appellate decision. Treat $135.5 million as what the jury found, not what was paid. Full post-trial history.
McKiver v. Murphy-Brown, LLC, E.D.N.C.; 980 F.3d 937 (4th Cir. Nov. 19, 2020)	Odor, flies, and truck noise from an industrial hog operation (Smithfield's hog production subsidiary); nuisance	Jury verdict, spring 2018: $75,000 compensatory to each of ten neighbors plus punitive damages; reported total $50.75 million	The district court applied North Carolina's punitive cap (the greater of $250,000 or three times compensatory, per plaintiff), cutting punitive damages to $2.5 million and the judgment to about $3.25 million. The Fourth Circuit affirmed liability for both compensatory and punitive damages but vacated the punitive amount and remanded for a rehearing without the parent company's financial evidence.
The five North Carolina hog trials, 2018 to 2019	Same defendant, same theory, 26 suits for roughly 500 neighbors	Juries awarded $549,902,400 across five trials; the punitive cap reduced that to $97,880,000. The third trial (Artis) is widely reported at $473.5 million gross, reduced to roughly $94 million by the cap; we could not fetch a primary source for the Artis figures this session, so treat them as reported.	Smithfield settled all remaining cases in November 2020 on undisclosed terms after losing the McKiver appeal. The gross verdicts were never paid as awarded.
Kamuda v. Sterigenics U.S., LLC, Cook County, Ill., No. 2018 L 010475	Ethylene oxide emissions from a medical sterilization plant in Willowbrook, Illinois; one plaintiff's breast cancer	Jury verdict, September 19, 2022: $363 million ($38 million compensatory, $325 million punitive)	A personal injury verdict for one plaintiff, not a property case. A later settlement resolved the claims of nearly 900 residents for a reported $408 million, which averages under $500,000 per claimant before fees.
Baker v. Saint-Gobain Performance Plastics Corp., N.D.N.Y.	PFOA contamination of Hoosick Falls, New York drinking water from plastics manufacturing	Class settlement of $65.25 million with Saint-Gobain, 3M, and Honeywell; final approval February 7, 2022	Allocated roughly $21 million to municipal and private well property classes, almost $8 million to a nuisance class, and nearly $23 million to a medical monitoring class serving more than 2,300 claims. DuPont did not participate. The property share is a third of the headline.
Doyle v. Fluor Corp. (Doe Run Herculaneum smelter), St. Louis Circuit Court	Lead emissions from a Missouri smelter contaminated a town's yards and depressed home values	$55 million class settlement approved April 4, 2012	Allocation: $26.4 million for annoyance and discomfort, $13.2 million for property damage, $13.75 million in attorneys' fees, plus costs and incentive awards. Valued against 453 class properties plus former owners. Paid in four annual installments. A separate personal injury trial over the same smelter produced a $358.5 million verdict in 2011, a different case with different plaintiffs.
Buck v. Republic Services, Inc., E.D. Mo., No. 4:13CV00801	Noxious odors from an underground fire at the Bridgeton Landfill; lost property value and public nuisance	Class settlement of nearly $7 million, preliminary approval April 17, 2014	More than 400 residences, claim form required, eligibility tied to living near the landfill between November 2010 and December 2013. A per-household figure in the low five figures, before fees.
Meeker v. Bullseye Glass Co., Multnomah County, Ore., No. 16CV07002	Heavy metal air emissions from a Portland art glass factory	$6.5 million fund plus a $1 million neighborhood air monitoring program, with a guarantee that the company use approved filters on all furnaces; final approval May 10, 2019	The class was defined by an expert's modeled emissions plume map, not a radius. The non-cash terms (filters, monitoring) are a large part of what neighbors got.
Aruba Petroleum, Inc. v. Parr, Dallas County Court at Law No. 5; Tex. App. Dallas, No. 05-14-01285-CV	Air emissions, light, and noise from natural gas wells near a family's 40 acre property in Wise County, Texas; intentional private nuisance	Jury verdict: $2.65 million for past and future physical pain and mental anguish plus $275,000 for lost market value, a total of $2.925 million for one family	Reversed and rendered on February 1, 2017. The Court of Appeals held there was no legally sufficient evidence that the operator intended the interference, and the family took nothing. A verdict is not a judgment until the appeals are over.
Martin v. KCBX Terminals Co., N.D. Ill., No. 1:13-cv-08376	Petroleum coke dust blowing from storage piles onto Chicago's Southeast Side	$950,000 class settlement with KCBX Terminals, Koch Carbon, and DTE Chicago Fuels Terminal; preliminary approval March 10, 2016, case closed July 2016	The smallest figure on this list, for a dust nuisance across a dense urban neighborhood. The bigger result was regulatory: the city's ordinance changes that forced the piles to be enclosed or removed.

Three patterns hold across the list. First, the cases that produced the largest per-household numbers were the ones with the most concrete, measurable harm to a specific property: a lake filled with sediment, a well full of PFOA, a yard that had to be dug up. Second, the cases that produced the largest headline numbers were punitive, and punitive awards are the part most likely to be cut by a statute, an appellate court, or a settlement. Third, class settlements for diffuse harms like odor and dust tend to land in the low five figures per household, and often the non-cash terms, the filters, the monitoring, the enclosure of the piles, were the part that changed daily life.

Anyone searching for what an environmental neighbor case is worth will run into figures that are accurate and still misleading. The traps repeat, and knowing them is what separates a citable number from a marketing number.

The total fund is not one firm's result, and it is not one plaintiff's result. A law firm's results page will say "$408 million" or "$65 million" without saying how many claimants shared it. The Sterigenics settlement covered nearly 900 residents. The Hoosick Falls settlement covered more than 2,300 claims across three separate classes. The Doe Run settlement paid $13.75 million of its $55 million to the lawyers before a homeowner saw a dollar. Divide before you compare. Two firms involved in the same Sterigenics litigation publish different totals, $408 million and $456 million, because each counts a different bundle of settlements; neither figure is wrong, and neither tells you what one household received.

Gross verdicts are not judgments, and judgments are not payments. The North Carolina hog juries awarded about $550 million; the state's punitive damages statute cut that to about $98 million before any appeal, and the Fourth Circuit then vacated the punitive award in the lead case for a redo. Courts also reduce verdicts through remittitur, and defendants settle for less than a judgment to end the appeal. The Parr family's $2.925 million became zero. Ask, for every verdict you read about, whether it survived post-trial motions and appeal.

Punitive damages are the part that disappears. Of the $135.5 million H&L Farms verdict, $125 million was punitive. Of the $363 million Sterigenics verdict, $325 million was punitive. Many states cap punitive awards by statute (North Carolina's cap is the greater of $250,000 or three times compensatory damages per plaintiff), and federal due process limits keep ratios in check even where no statute applies. When you see a nine-figure nuisance verdict, find the compensatory number; that is the part that reflects the harm to the property and the people.

Some of the biggest figures never went to neighbors at all. The 3M AFFF settlement, often cited as the largest PFAS resolution in history, is between $10.5 billion and $12.5 billion under the settlement agreement 3M filed with the SEC, and the class is public water systems, not households. Flint's water crisis settlement, widely reported at $626 million, was a fund shared among thousands of claimants and many law firms; we could not fetch a primary source for that figure this session, so treat it as reported. And regulatory penalties are not recoveries: the California Public Utilities Commission's $1.6 billion San Bruno pipeline penalty against PG&E in April 2015 was a $300 million fine to the state's general fund, a $400 million credit spread across gas customers' bills, and $850 million in shareholder-funded safety work. Not one dollar of it was damages to a neighbor.

Personal injury and property damage are different cases with different numbers. The $358.5 million Doe Run verdict was a personal injury trial for lead-poisoned plaintiffs; the $55 million was the property class. The Sterigenics verdict was one woman's cancer case. A data center neighbor whose complaint is noise, dust, or a dry well is in the property and nuisance column, where the analogous per-household figures are far smaller than the headlines.

Strip the caveats away and the analogous record points to a consistent set of factors. These are the things a lawyer will ask about first, and they are the things you can start documenting today.

• Proximity and a defined footprint. Every class on this list was drawn around the source: a plume map in Portland, the neighborhoods around the Bridgeton landfill, the well-dependent area of a groundwater basin in Oregon. The 2026 data center noise classes are converging on a radius of about one mile. Being inside the line, and being able to show why the line is where it is, is the threshold question.
• Measurable harm. Sediment in a lake, PFOA in a well, lead in a yard. For data centers, the equivalents are decibel readings at your property line compared against the local code, before and after well tests, and dated photographs of construction runoff. The Sterigenics and Hoosick Falls cases were built on emissions and water data that existed before the suits were filed. Our evidence guide covers how to build that record.
• Property value evidence. The Doe Run settlement was valued against the uncontaminated value of 453 specific properties; Hoosick Falls set aside a separate property diminution fund; the Parr jury put a discrete $275,000 on lost market value. Appraisals, listing histories, and buyer feedback that names the facility turn a general grievance into a number. See our guide on data centers and property values.
• The number of households. More affected homes mean a class is viable, a settlement fund is larger, and a defendant has more reason to resolve the whole problem. The trade-off is that a fund divided among 900 or 2,300 claimants pays each one far less than the headline. Our class action guide explains how the math works.
• Punitive exposure. The largest verdicts here came where juries found an operator knew and did not act. The Fourth Circuit pointed to abundant evidence of Murphy-Brown's conscious disregard of conditions at the farm; the H&L Farms jury put $125 million on the construction defendants' conduct. For a data center, the analog is a paper trail: complaints logged and ignored, a noise or stormwater permit condition violated, a promised fix that never came. Keep every email and every meeting minute, because the operator's response to your complaints is evidence.
• The remedy you ask for. An injunction fixes the problem; damages compensate for it. They are not exclusive, but a case pleaded for one can leave the other behind. The original Granbury, Texas state court suit over the Wolf Hollow bitcoin mine seeks a permanent injunction against the fan noise rather than money, which is exactly the remedy operators fear most and also the outcome that leaves years of lost sleep uncompensated unless damages are claimed too. The Bullseye Glass settlement shows the middle path: cash, plus filters, plus monitoring.

Two more factors cut against value, and the record is honest about them. Deadlines are short and unforgiving, and a nuisance that has been tolerated for years is harder to present than one documented from the first week; our deadlines guide explains why. And the state's law on intent, caps, and the measure of nuisance damages can change everything: the Parr family lost on Texas's intentional nuisance standard, and North Carolina's cap took away 82 percent of what the hog juries awarded.

The first noise class to settle, or the first construction well case to reach a verdict, will be reported as a single number. When it lands, ask the questions this guide asks of every figure above: how many households share it, how much is punitive, what non-cash terms are attached, whether the operator admitted anything, and whether the number is a verdict, a judgment, or a payment. The Amazon settlement already teaches the pattern: $20.5 million sounds like a windfall until you learn it funds wells and water lines for a basin serving tens of thousands of people and that the defendant admitted nothing.

For your own situation, the useful comparison is not the headline but the category. If your complaint is a dry or contaminated well, the Oregon, Hoosick Falls, and Herculaneum cases are the ones to study, and the lesson is that baseline testing and a defined footprint decide the case. If it is noise, no analogous verdict exists yet in this field, and the value will be built from decibel readings, the local code, and the number of neighbors who can document the same thing. If it is construction runoff or damage, H&L Farms shows what a jury will do with photographs of a destroyed property and evidence that warnings were ignored. Our guide on what a data center lawsuit is worth walks through each category of recovery.

Whether your facts fit a claim, and in which category, is a question for a lawyer. Our free case review connects you with an independent attorney in our network who handles data center cases in your state, at no cost and with no obligation. We update this page as settlements and verdicts are confirmed; the date at the top tells you how current the record is.