Batch 21 — Psychiatric labels and fitness referrals applied to safety-relevant engineers

Date of work: 7 Oct 2026. Extends Batches 1–20. Nothing earlier is deleted. Carry-forward corrections stand (Diaz-Robainas: retaliatory evaluation order, no examination occurred; Petitt: pilot, comparison only, never counted as an engineer; Petitt settlement terms confidential; Diaz-Robainas NRC penalty paid 3 Dec 1996).

0. What this batch did, in one paragraph

It went back to the engineering-specific leads, read the Diaz-Robainas primary decision in full, found and read a second engineering case not previously in the archive (Patrickson v. Entergy), and tested other leads. Result: two adjudicated engineering episodes, both US nuclear (1991, 2003). In both, the instrument was a fitness-for-duty referral, not a diagnosis. No case was found of an identified clinician diagnosing a safety-reporting engineer with a psychotic or personality disorder. That is a "not found in what was searched", not a statement that it never happened. Search was heavily weighted to US Labor Department records; international and local-language coverage is a gap (Section 6).

1. Definitions used (from the brief)

Label types: L1 diagnosis by identified clinician; L2 employer/colleague informal label; L3 fitness-for-duty referral or demand without diagnosis; L4 hospitalization/detention; L5 label alleged by the person, not corroborated.

Instrument categories A–G carry over from Batch 19 (A = employer fitness-for-duty or contractual evaluation).

Every row also carries: engineering role verified? / fate of the hazard / Griffin test (what unusual behaviour was observed before the referral).

2. Plain-language account, chronological (what the records support)

- By 1991: US nuclear licensees run mandatory fitness-for-duty programs (10 CFR Part 26, cited in the decisions; start date not verified here). The same programs that exist to catch impairment give managers a discretionary route to order a psychological evaluation. The Labor Secretary later said NRC mandates "do not prevent" an employer from abusing such a policy (paraphrase of Diaz-Robainas, 1996).

- 1991, Florida (Turkey Point): Senior I&C engineer Regino "Richard" Diaz-Robainas, in dispute with managers over pressure-transmitter replacement and other projects, is ordered to a psychological fitness-for-duty evaluation on 30 July 1991. He does not attend. He is fired 19 August 1991 for refusing. The Secretary of Labor (19 Jan 1996) finds the order itself was motivated solely by retaliation. No examination ever took place.

- 1993–1996: The doctrinal question is contested inside the Labor Department. The ALJ (29 Oct 1993) treated the order as non-punitive and not adverse. The Secretary (1996) reversed: an order to submit to evaluation can be an adverse action, and motive is the issue.

- 2003, New York (FitzPatrick): System engineer Carl Patrickson, a self-described "prolific reporter of problems" who had told the NRC about a pump-room ventilation concern in 1997 and OSHA about other hazards in 2003, has a hallway confrontation with the plant's general manager and sends an all-capitals email. On 27 March 2003, eight days after OSHA signed its citation and three days after he showed those findings to HR, site management and corporate headquarters decide on a for-cause drug/alcohol test and a psychological evaluation. Drug test negative. The psychologist's report to the employer found "no significant pathologies" but recommended counseling and six months of intensified behavioural observation. He is out 31 days, returns 28 April on those conditions, and is fired 20 Nov 2003 after a performance plan. ALJ recommended decision 3 Mar 2005; ARB final decision 31 Aug 2007. Per the 7 Oct engineering addendum in the uploaded Complete library (not independently verified by me): the ALJ's recommendation was favourable to Patrickson, the ARB rejected the retaliation claim and reversed reinstatement, and the Second Circuit upheld the result on 18 Dec 2008. He is NOT counted as an established retaliation finding.

- 2007: ARB adopts the "materially adverse" standard across the statutes it hears (Powers, 31 Aug 2007), the test later used to decide whether a referral counts.

- 2022: Petitt (pilot) ARB decision treats a compulsory psychiatric evaluation as adverse only case by case (carried; comparison).

3. Case cards

3.1 Diaz-Robainas v. Florida Power & Light, 92-ERA-10 — VERIFIED (primary read in full this batch)

Primary: Sec'y of Labor, Decision and Remand Order, 19 Jan 1996: https://kkc.com/wp-content/uploads/2023/02/D-Rv.FPL_.pdf (scanned copy; verify pagination against the reporter). NRC EA 96-051: https://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/reactors/ea96051

1. Person/role: Regino R. Diaz-Robainas (NRC and coworkers use "Richard"). Engineer at Florida Power 1980–19 Aug 1991; Lead Engineer, Instrument & Control (1985); Senior Engineer when fired. Turkey Point Nuclear Plant. Engineering role VERIFIED from the decision's own findings.

2. Concern: disputes with supervisor Bob Wade over replacing pressure transmitters that were to sense a loss-of-cooling accident, and over the Westinghouse setpoint study and ERDADS (a post-TMI operator-information system).

3. Label: none clinical. Instrument = order to a psychological fitness-for-duty evaluation with the employer's consultant Dr. Dennis Johnson (L3). Informal characterisations in the record (L2): Hosmer saw him as "overwhelmed and stressed out"; Wade's appraisal called him not "practical at times" and unwilling to entertain others' opinions.

4. Initiator and stated grounds: John Hosmer, Director of Nuclear Engineering. Grounds shifted: stress and declining performance, then fear of "sabotage". Early July 1991 the division president asked managers about employees in performance counselling who might attract "adverse newspaper or NRC reactions" and suggested considering fitness for duty. Hosmer acted on 26–30 July after Robainas threatened to go to the Miami Herald.

5. Sequence: Mar 1990 drug-policy objection; Feb 1991 below-average appraisal and 23 Feb letter alleging retaliation; 30 Apr interim review; 2 May internal Speakout complaint; 30 Jul order; 2 Aug no-show, badge pulled; 9 Aug Hosmer learns of NRC contact; 19 Aug refusal and firing; complaint filed 29 Aug 1991.

6. Independent assessment and contrary evidence: first-line supervisor Pagnozzi saw him daily and never questioned his fitness; coworkers and the ombudsman saw no instability; Dr. Johnson never met him and had only partial documents. Contrary: the Secretary accepted that the Feb and Apr performance ratings had valid reasons and were not shown pretextual, accepted Robainas had said he was under stress, and agreed Hosmer had a duty to ensure fitness. The ALJ had treated the evaluation as non-punitive.

7. Findings: ALJ R.D.&O. 29 Oct 1993: dismissed. Secretary 19 Jan 1996: reversed on discharge; "Hosmer's explanation is a pretext"; order "based solely on retaliatory animus" (quoted by NRC from the Decision at 8). Remanded for remedy; not a final remedy decision. NRC: $100,000 civil penalty (paid 3 Dec 1996, per DD-97-20, carried). Settled before ALJ remedy ruling (carried).

8. Consequences and hazard: fired; reinstatement and back pay ordered; settled. Fate of hazard (UPGRADED from LEAD): the Secretary records that Florida Power "ultimately agreed" and replaced the transmitters, and that the plant eventually agreed with him on ERDADS. The record does not show whether he was credited.

9. Griffin test: first-line observer saw nothing; the referral decision-maker had seen him twice in 1991. Result: referral not supported by observed behaviour (Secretary's finding).

10. Remaining uncertainty: settlement terms; whether the setpoint study concern was resolved; the ALJ's full reasoning (not read). Next source: ALJ R.D.&O. 29 Oct 1993.

Status: VERIFIED.

3.2 Patrickson v. Entergy Nuclear Operations, 2003-ERA-22 — facts VERIFIED from the ALJ record; final outcome: retaliation NOT established (ARB reversed ALJ; Second Circuit affirmance per addendum, unverified by me)

Primary read: ALJ Recommended Decision and Order, 3 Mar 2005, pp. 1–about 31 of the file: https://www.citizen.org/wp-content/uploads/entergydiscriminationsuit.pdf. ARB final decision, ARB Nos. 05-069 and 05-070, 31 Aug 2007: NOT read (oalj.dol.gov blocks automated access); only DOL digest summaries seen: https://www.dol.gov/agencies/oalj/PUBLIC/ARB/REFERENCES/CASELISTS/08_2007

1. Person/role: Carl R. Patrickson, engineer, James A. FitzPatrick plant (NY Power Authority, then Entergy from Nov 2000). Maintenance engineer 1989; field engineer 1991–2000 (monitoring plant modifications); system engineering "assistant engineer" from May 2000, monitoring systems for proper operation and maintenance. Engineering role VERIFIED. Safety responsibility is system monitoring and problem reporting; he is not a titled safety engineer.

2. Concern: emergency service water pump room ventilation: closed fire dampers could let pump motors overheat and fail in a fire. Reported to NRC 1997 (RI-1997-A-0126) and again 31 Mar 2003 (RI-2003-A-0053); also six OSHA-type items reported to the ethics line (Jun 2002) and OSHA (13 Jan 2003). OSHA inspected in Mar 2003.

3. Label: L3 plus L2. For-cause drug/alcohol test and psychological evaluation via the employer's EAP vendor (EMAX). Employer's descriptors: "aberrant behavior", "out of character", "excited emotional state". The general manager testified: "I wouldn't say screaming, you know, crazy". His worst-case rationale: someone "acting crazed" might injure someone or tamper with the plant. No diagnosis. Clinical output (L1-adjacent, but not a diagnosis): the EMAX report found "no significant pathologies", recommended short-term stress counselling and at least six months' intensified behavioural observation; the report was written by the EMAX vice-president, but a local psychologist, Dr. Joel Richman, saw him.

4. Initiator and grounds: general manager O'Grady (email 26 Mar), then a 27 Mar meeting of the HR director Zimmerman, O'Grady, engineering director Limpias and site VP Sullivan, plus a call to Entergy Nuclear Northeast in White Plains. Cited policy: AP 11.01 fitness-for-duty program, derived from 10 CFR Part 26. Sullivan said he feared Patrickson "could do harm to himself" or equipment, and admitted he knew of no past harm to equipment.

5. Sequence: 1996–97 disputed evaluations and 1997 NRC report; Jun 2002 ethics-line report; 13 Jan 2003 OSHA letter report; 15 Jan and 25 Feb random tests; 19 Mar OSHA citation signed (minor fine on one item); 24 Mar he shows HR the OSHA findings; 25/26 Mar hallway encounter; 27 Mar email and for-cause referral; leave; 31 Mar second NRC report; 1–3 Apr negative drug test, evaluation, report; 22 Apr his OSHA retaliation complaint; 28 Apr return under agreement; 2003 performance plan; 20 Nov termination.

6. Contrary and independent evidence: OSHA's regional administrator (17 Jul 2003) found referral legitimate because of "out of the ordinary behavior". The HR director described him as quiet and calm, which supports both "out of character" and "no history of instability". Fewer than five for-cause tests since 2000 at the plant. Patrickson testified he knew no one else kept on 31 days' leave after a negative test or put under six months' observation; Zimmerman said it was the first such programme Entergy had heard of. Patrickson told HR he "kind of used the NRC thing as a shield" for a year or two (Zimmerman's memo; Patrickson called "spiteful" too strong). O'Grady admitted knowing of the OSHA complaint but testified the ventilation issue was never brought to him; Patrickson said his 2002 "Short List" (which included ventilation) was emailed to O'Grady, though he admitted never discussing it with him. Unresolved conflict about who knew what.

7. Findings: OSHA 17 Jul 2003: no violation. ALJ 3 Mar 2005: finding on the evaluation NOT read. DOL's own ARB digest says the ALJ found the performance-plan meetings were adverse action and found disparate treatment, and that the ARB reversed both and found no pretext for the discharge. That implies the discharge and plan claims were not sustained; it does not say how the evaluation referral was decided.

8. Consequences and hazard: 31 days' paid leave, return-to-work conditions (stress counselling, six months' observation), then discharge on 20 Nov 2003. Fate of hazard: NRC wrote (28 Apr 2003) that absent new information it had no basis to revise its earlier conclusion that the problems were valid but adequately addressed. Patrickson testified permanent modifications were cancelled. ALJ record shows a 2003 mod list marking two as cancelled. NOT resolved here; no independent check of the 2003 fire analysis.

9. Griffin test: observed behaviour = a raised-voice hallway exchange and an all-caps email, in a man others call calm. The test is met on the employer's account and contested on his. The independent clinician found nothing pathological. Neither fact settles motive.

10. Uncertainty / next source: the ARB decision text (05_069.ERAP; try mirror or Westlaw); ALJ conclusions section; whether the evaluation was held adverse; later NRC discrimination findings under allegation RI-2003-A-0053.

Status: VERIFIED for facts quoted from the ALJ record. Final adjudicated outcome is against the retaliation claim; whether any reviewing body specifically ruled on the evaluation referral remains unread. The case stays in the register as a contrary/qualified row: independent clinician found no pathology, yet conditions were imposed, and reviewers did not find retaliation.

4. Case table

5. Analysis

Labels (what actually appears). In the engineering-specific records the dominant label is "fitness for duty" (L3), accompanied by informal terms (L2): "stress", "aberrant", "out of character", fears of sabotage or self-harm. The words paranoid, delusional, persecution complex and personality disorder do not appear in either verified engineering record. They do appear in adjacent cases (Griffin, Petitt) and in the NTSB airman lead. I have no evidence these terms were applied to engineers; I also did not search enough to say they were not.

Mechanisms (carry the nine from Batch 19). Seen here: 1 credibility displacement (the order replaces argument about transmitters or fire dampers with argument about the man); 2 removal at moment of speech (Diaz-Robainas: after the press threat; Patrickson: 8 days after OSHA signed its citation); 3 procedural shadow (31-day leave, return-to-work conditions that a negative result did not end); 7 welfare/safety framing as pretext (Secretary found it so in 1996; contested in 2003). Mechanism 4 (durable record) is plausible in Patrickson, with six months' documented observation, but no document shows it was used later.

What the two cases do not show: that the hazard was wrong. In the one case with a fate on record, the engineer's technical position was later adopted. In the other the regulator found the concern had been handled. Adjudication here decides the employment question, not the safety question.

Differences by period. The legal treatment shifts: ALJ 1993 (non-punitive, not adverse) → Secretary 1996 (can be adverse; motive decides) → ARB 2007 (materially-adverse standard) → ARB 2022 (Petitt: case by case). Practice shifts: the 1991 order was blunt (refuse and you are fired); the 2003 process was institutional (HR, corporate, EAP vendor, return-to-work agreement). That is two data points, not a trend.

Differences by industry. Only nuclear (with its mandatory fitness-for-duty regime) is documented. Aviation (Petitt), trucking (Griffin, Frausto, per digest: "bizarre and disruptive behavior" upheld as a legitimate ground), and rail have their own safe-harbour rules. Whether engineers outside regulated-FFD industries (civil, structural, chemical, product) experienced this is not answered: no case found; search coverage thin.

Differences by country. Not established. Only US records were substantively reached. See gaps.

What the evidence can say about recurrence. Two adjudicated engineering episodes (1991, 2003) in one industry in one country is a floor, not a rate. Biases: only contested, adjudicated or published matters surface (Diaz-Robainas and Petitt both settled before remedies were finalised); the two cases I could read are ones where the employee kept litigating; nuclear fitness referrals are routine, so a referral alone is weak evidence of motive; engineers who accepted an evaluation quietly leave no record; databases code these as discharge, discrimination or access authorization rather than psychiatric.

What would be needed to estimate frequency: (1) a census numerator: all OSHA/ALJ/ARB/DOE Part 708 files mentioning psychological, psychiatric, fitness-for-duty or EAP, coded by role; (2) a denominator: allegation counts (NRC allegation program; DOE employee-concerns offices) or headcount of reporters; (3) a comparison group: referral rate among non-reporters (NRC fitness-for-duty semiannual reports list for-cause "behavior" tests per plant, e.g. Waterford 3, Jul–Dec 2002: one; https://www.nrc.gov/docs/ML0306/ML030640743.pdf); (4) outcomes: how many were found retaliatory, upheld, cleared. Without a comparison group, no rate can be read as elevated.

6. Search log (Batch 21 only)

Repositories: DOL OALJ Nuclear and Environmental Whistleblower Digest Div. XIII (fetched; page truncated before XIII B 11 "Referral to employee assistance program/medical or psychological evaluation", the key section, NOT read); DOL ARB caselists (Aug 2007; ERA lists); NRC enforcement (EA 96-051; ML20069K254 snippet); NRC FFD six-month reports (snippets); DOE Office of Hearings and Appeals summaries (Parts 708 and 710; snippets only); Public Citizen-hosted ALJ decision; kkc.com and whistleblowers.org mirrors of DOL decisions; NTSB; news (Boeing); Substack (one item).

Terms: engineer + safety + psychiatric/psychological evaluation/fitness for duty/paranoid/mentally unstable/delusional; "Patrickson Entergy"; "Wells Kansas Gas"; "Smith v. Esicorp"; Part 708/710 + psychologist + safety concerns; Russian: engineer + psychiatric expertise + unsafe + compulsory hospitalisation.

Languages: English; Russian (one query; irrelevant results; poor query design, not evidence of absence). NOT searched: Japanese, Korean, French, German, Spanish, Chinese, Ukrainian beyond incidental, UK tribunal databases, Canada, Australia.

Access gaps: oalj.dol.gov decision pages are robots-blocked (Patrickson ARB, Griffin, Mandreger unread); Westlaw/Lexis/ADAMS full text; DOE OHA index not queried directly; no professional-archive, union, oral-history or legislative-hearing search this batch.

Reviewed, no psychiatric element seen, not counted: von Bargen (Sandia safety engineer, TBH-0034); Trask (LANL engineer, WBU-15-0003); Hasan (various ERA civil/structural engineer cases); Tracanna (TAPS electrical inspector; constructive discharge rejected); Hoffman v. NextEra (assistant operations manager; constructive discharge rejected); McNeill v. Crane Nuclear; Boeing: Woods (disclosed his own ADD/OCD/depression at hire; no employer labelling seen), Barnett (family alleges PTSD from the work environment, which is alleged harm, not a label applied), Salehpour, Fauver.

Comparison section (not engineers, not counted): Petitt (pilot); Griffin and Frausto (drivers; behaviour grounds upheld, per digest); NTSB Order EA-5695, 2 Jan 2014 (airman; FAA flight surgeon sought psychiatric evaluation citing "suspected paranoid ideation" in 1970s records and "highly improbable" allegations in a 2012 letter; outcome unread; respondent not recorded); Indian Point security officer (S.D.N.Y., 290 F. Supp. 3d 234; for-cause evaluation after a reported hallucination; not a safety report; outcome unread). Unverified current item: a Substack account of a Pennsylvania involuntary commitment mentions whistleblower litigation and a record using "paranoid and delusional"; engineering role not established; single advocacy source; not counted.

7. Next batch (22) queue

1. Re-fetch DOL Digest Div. XIII with a much higher size limit and read XIII B 11 in full; list every case and classify by role.

2. Get Patrickson ARB text (mirror/Westlaw) and the ALJ conclusions pages; then NRC allegation RI-2003-A-0053 follow-up.

3. Griffin and Mandreger primaries via mirrors (kkc.com, whistleblowers.org, CourtListener); Wells 85-ERA-22; Floyd; Freels; Smith v. Esicorp; Saporito remand.

4. DOE Part 710 channel: query OHA for engineers with a safety-report history who were sent for a DOE psychologist.

5. NTSB EA-5695 full text (comparison).

6. International: UK employment tribunals (protected disclosure + psychiatric/fitness), Japan, Korea (Gori, Kori), Canada (CNSC), Soviet-era sources (Podrabinek, Bloch & Reddaway, Memorial lists), in local languages with better query design.

7. Professional and union archives (NSPE Board of Ethical Review, IEEE, ASME, IBEW, UCS, GAP); oral histories; Senate Hanford hearing (S. Hrg. 113-370) read for any psychiatric references.

8. Frequency design: pull NRC FFD six-month reports for 3–5 plants over 10 years; cross-match against allegation data via FOIA.

Standing rules unchanged: rebuttals first; Griffin test on every referral; missing field = opening question; version for every statute row; "psychiatric" is a trailing term.