EPA lets Pavilion8 PEMS stand in for NOx analyzers in Arkansas

EPA found that trained Pavilion8 models matched hardware NOx analyzers on seven Arkansas turbines, with relative accuracy of 1.38% to 3.15%. It approved them in their place.

  • Not rated an organization

    EPA approved the models on a demonstration against the analyzers, and the letter reports no saving or other outcome for the cooperative.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Intended benefit.
Evidence

foundstudiedreportedclaimed

An independent regulatory document supports the benefit.

Compared with

Certified hardware NOx analyzers (CEMS) on the same turbines. Measured.

The system’s part
Each model's hourly NOx predictions were compared with the hardware analyzer on the same turbine, so the accuracy figures are the models' own. EPA writes that each PEMS was trained on operating data and that a model for each turbine was developed from the training data.
Limits
Not yet checked.
Costs and harms
Not yet checked.
Later
Not yet checked.

Download data Cite Suggest a correction

Benefit type
Improved Task Accuracy
System
Pavilion8 Software PEMS (predictive emission monitoring systems) on Units G1 through G7
Made by
Rockwell Automation (maker of Pavilion8, which the letter names only as a product)
Used by
Arkansas Electric Cooperative Corporation (AECC), Harry L. Oswald Generating Station
Where
Pulaski County, Arkansas, United States
Stage
In a pilot
Date
, when the source was published
Kind
Aggregate result
Broad benefit
Estimating the Quantifiable Characteristics of Products, Events, or Information, Monitoring Processes, Materials, or Surroundings
Collections
Scrutiny and findings
Last revised
31 October 2026

Sources

  1. Regulatory document, by authors with no tie to the system.

    U.S. Environmental Protection Agency, Clean Air Markets Division. Letter to the Arkansas Electric Cooperative Corporation approving alternative monitoring systems (PEMS) for Units G1 through G7 at the Harry L. Oswald Generating Station (ORISPL 55221), signed by Reid Harvey, Director, 21 November 2012.

    Tables 1 through 4 also summarize the results of AECC's statistical analyses of the paired PEMS and CEMS data, which show that the NOx lb/mmBtu output from each PEMS passed all three required statistical tests.
    A model for each combustion turbine was then developed from the training data.

Details

AI system

Workflow

Each PEMS was trained on sensor readings of turbine operation while NOx was measured at the same time. Each model predicted hourly NOx emission rates from inputs such as gas flow, steam injection and humidity.

Follow-up

Conditions for replication

Input values stay within the operating envelope of the training data. Daily checks, monthly audits and periodic accuracy tests continue, with retraining when operations change.

Entry details

Sector

Air emissions monitoring (power generation).

Purpose

Rockwell Automation built the Pavilion8 software PEMS to predict a combustion turbine's NOx emissions from its operating data.

Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this record

Peter H. Vartanian. (2026). EPA lets Pavilion8 PEMS stand in for NOx analyzers in Arkansas (AIBD-0910). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0910.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)

BibTeX Data Field definitions