Plant data assessment — TEP-Plant-10days

Tennessee Eastman benchmark — AI and AO · period analysed 2026-04-01 00:00 to 2026-04-10 23:57 · 52 tags · prepared by Yahya Nazer, CTRL Designer LLC · 11 September 2026

1. What this is

This is a data-driven assessment of Tennessee Eastman benchmark — AI and AO. It reads the plant's own historian export and asks three questions: is the data good enough to trust, when did the plant leave normal operation, and which of those departures are worth acting on. Nothing here comes from a model of how the plant should behave. Everything is measured against a window of the plant's own history chosen as normal.

The headline. 3 cause(s) came out worth acting on and 1 were set aside after review, from 52 tags sampled every 3 minutes over 2026-04-01 00:00 to 2026-04-10 23:57. Normal was learned from 2026-04-01 00:00 to 2026-04-02 23:57 (47.95 h). 7 question(s) still need the plant to answer.

The work runs in five steps, and each one writes its own report so any number can be traced back to the raw export:

  1. Prepare — the raw export becomes one clean, regular table. Every repair is recorded.
  2. Explore — the tags are grouped by how they move together, and the scope is chosen: which tags, which period, and which window counts as normal.
  3. Analyze — a model of normal is built from that window, then the whole period is scored against it.
  4. Review — every finding is challenged before anyone acts on it, and given a verdict.
  5. Report — what survives becomes the assessment, with the actions and the open questions.

Charts are live: drag to zoom, double-click to reset, click a name in the legend to hide it. Every report also carries Last and Next buttons at the bottom, so you can read straight through without coming back here.

2. The reports

StepReportWhat is in itSize
AStep A — Overall
A-Data/TEP-Plant-10days-AIAO/overall_report.html		What the study is, the data it started from, what was done and what came out. Start here.11 KB
AStep A — Preparation
A-Data/TEP-Plant-10days-AIAO/prepare_report.html		Every tag's data quality: what was missing, frozen, spiking or arriving late, and what was kept.28 KB
AStep A — Bad tags
A-Data/TEP-Plant-10days-AIAO/bad_tags_report.html		One page per flagged tag, with the trend and a zoom on each problem zone.56 KB
AStep A — Fixed tags
A-Data/TEP-Plant-10days-AIAO/fixed_tags_report.html		The same tags as received, then the bad zones, then after repair - the three rows, so you can judge every correction.47 KB
BStep B — Explore
B-Scope/TEP-Plant-10days-AIAO/explore_report.html		The tags grouped by how they move together, and the scope: which tags, which period, and which window counts as normal.23 KB
CStep C — Analyze
C-Analysis/TEP-Plant-10days-AIAO/analysis_report.html		The PCA model of normal, when the plant left it, which loops cycle, and the ranked findings.38 KB
DStep D — Review
D-Review/TEP-Plant-10days-AIAO/review_report.html		Every finding challenged and given a verdict, then grouped into causes with a chart each.34 KB
EStep E — Report
E-Report/TEP-Plant-10days-AIAO/report.html		The assessment: what to act on, what was set aside and why, the actions, and the questions for the plant.18 KB

Total 76 MB including all charts and tables.

3. Questions and comments

Every finding in this study is a question until the plant answers it. If something looks wrong, or a window matches work you know about, that is exactly the feedback that makes the next pass better.

When you write, it helps to say:

Email info@ctrlDesigner.com

Yahya Nazer · CTRL Designer LLC

© 2026 CTRL Designer LLC. All rights reserved.
Yahya Nazer · info@ctrlDesigner.com