He Submitted My 4-Year Superfund Contamination Protocol to the EPA — Then the Hearing Required the Licensed Site Assessor of Record
He Submitted My 4-Year Contamination Protocol to the EPA Under His Name — Then the Enforcement Hearing Required the LSA of Record
Dr. Camille Osei was at drill location 14 at the Blackwood Superfund site on a Tuesday in September when Marcus handed her the next sample label.
She took it.
She read it: BKW-14-S2-VAPOR.
She said: “Before we pull from 14, I need you to move the equipment to the 15 position and set up. We’ll come back.”
Marcus said: “Why does 14 come before 15 in the sequence but we’re going to 15 first to set up?”
She said: “Because the groundwater flow at this site pulls southwest at approximately 1.4 meters per day. If I draw the vapor sample from 14 before I have everything set up at 15, there’s a risk the mechanical disturbance from the drill equipment at 15 affects the subsurface pressure differential and contaminates the 14 draw. I take 14 clean, then I take 15 downstream.”
Marcus said: “So the order matters that specifically.”
She said: “The order is the science. A wrong sequence doesn’t just give you a bad sample. It gives you a believable bad sample. You can’t tell from looking at it that it’s wrong.”
He moved the equipment.
She opened the glass Dewar flask.
The flask was vacuum-insulated and 500mL capacity — standard for volatile organic vapor collection at PCB-contaminated sites.
The neck was wrapped in yellow evidence tape from the previous site visit, four weeks ago.
Beneath the tape, an adhesive label carried her LSA registration number: GA-4471.
She had been a Licensed Site Assessor since 2015.
She was the only LSA on the Blackwood project team.
She had been the only LSA on the project team for four years.
She positioned the draw tube at the correct depth.
She said: “Starting 14-S2 vapor draw.”
Marcus logged the time: 9:47 AM.
She pulled the sample.
She capped the flask.
She initialed the chain-of-custody form.
She signed the bottom: C. Osei, LSA #GA-4471.
Marcus signed as witness.
She carried the flask to the cooler.
It was the 147th sample she had collected at the Blackwood site.
It was a Tuesday.
The site was ordinary to her.
She had been coming to the Blackwood site since Year 1 of the remediation contract.
The Blackwood site was a former industrial transformer facility in rural South Carolina.
The transformers had contained PCB-based cooling oil.
Over decades, the oil had leached into the soil and was tracking southwest in the groundwater plume toward a small reservoir 2.4 kilometers from the facility.
The EPA had designated it a Superfund site in 2020.
Halcyon Environmental Group had won the remediation contract in 2021.
Grant Ellison, Senior Project Manager, had signed the contract.
Camille had designed the sampling protocol.
The protocol was not simple.
A PCB-contaminated groundwater plume required a sequenced approach: sample upflow locations first to establish an uncontaminated baseline, then sample the plume body in the direction of flow, then close with downstream confirmation draws.
If you sampled out of sequence, you didn’t know what you were looking at.
She had built the sequence from the site’s groundwater model, which she had run herself.
The model had taken eight weeks.
The sequence had been revised three times as new data came in.
It was now in its fourth iteration, updated most recently in April.
She had sent Grant each iteration.
He had forwarded them to the EPA project officer with cover notes signed: G. Ellison, Lead Investigator.
She had been on every site visit.
He had attended two.
The Before was a Thursday evening in Year 2.
The Year 2 EPA progress meeting had been that afternoon.
Grant had driven to the EPA Region 4 office in Atlanta.
Camille had been in the field.
She had sent him a 3-page memo the previous Sunday night: the southwest flow rationale, updated from the latest modeling run, explaining why the Year 2 sampling grid needed to shift 20 meters east.
The memo had data tables.
It had a map.
It had three paragraphs of technical explanation.
Grant had taken the memo to the EPA meeting.
He had presented the southwest flow rationale to EPA Project Officer James Cully.
He had quoted the memo’s technical language verbatim.
Cully had asked two clarifying questions.
Grant had answered them — one from memory, one by reading from the memo.
That evening, Grant had sent Camille a single-line email.
It said: “Good meeting. They liked the SW flow rationale.”
She had read it.
She had typed: “Thanks.”
She had closed her laptop.
She had not typed: “I wrote that rationale on Sunday night.”
She had not typed: “Which part of the rationale did they like?”
She had not typed anything else.
She had written “Thanks.”
She had closed the laptop.
The final Blackwood Method protocol document appeared on the EPA Superfund remediation database on a Wednesday in November.
She found it by searching the database on her lunch break.
She had known it was coming.
Grant had told her the submission was being finalized.
He had not shown her the document before it was submitted.
She read the cover page.
THE BLACKWOOD METHOD: A PROTOCOL FOR PCB-CONTAMINATED GROUNDWATER REMEDIATION AT POLYCHLORINATED BIPHENYL SUPERFUND SITES.
Developed by: G. Ellison, Lead Investigator, Halcyon Environmental Group.
She scrolled to the team page.
Project Team:
Grant Ellison, Senior Project Manager — Lead Investigator
Dr. Camille Osei — Field Sampling Lead
Marcus Webb — Field Technician
She read “Field Sampling Lead.”
She had a Dewar flask on her bench.
The evidence tape was still on the neck from the last site visit.
She took the flask from the bench.
She held it.
She set it on the corner of her desk.
She did not put it on the shelf.
She opened the project file.
She scrolled through the chain-of-custody forms.
There were 147.
Each one signed: C. Osei, LSA #GA-4471.
She closed the file.
She did not annotate it.
She went back to work.
(Drop “CAMILLE” in the comments if you want to see what happened when the EPA enforcement hearing required the LSA of record.) 👇
The Year 4 quarterly EPA meeting was a Thursday in January.
Grant drove to the EPA Region 4 office.
Camille sat in the back row.
She always sat in the back row at client and regulatory meetings.
She was there because Grant needed her for the technical questions.
She was in the back row because Grant was the Lead Investigator and it was his meeting.
He was at the table.
He had the sampling maps.
Her maps.
She had generated them in the GIS software on the third Tuesday of every month, updated with the latest monitoring data.
He had the maps.
He was using a laser pointer on the projection screen.
He was pointing to drill location 9.
EPA Project Officer James Cully said: “The vapor pressure correction at site 9 — I’m seeing a temperature adjustment factor that looks nonstandard. Can you explain the rationale?”
Grant said: “Sure. Our team applied a modified Henry’s Law adjustment for the site’s subsurface temperature. At 14 meters depth, the site runs about 2 degrees above the standard ambient reference temperature, which shifts the volatility coefficient. We compensated for that.”
He had read it in her field notes.
She had written the temperature correction in the month-3 site report.
She had explained it to him in a phone call in month 4 when he didn’t understand the reference temperature issue.
She had explained it again in month 8 in an email.
He had read the explanation three times across four years.
He had described it, just now, accurately.
He had described it in the first-person plural.
Cully was writing in his notes.
His notes said, under the heading “Ellison team technical response:” and then the explanation.
Cully said: “Makes sense. Dr. Osei — anything to add?”
She said: “The adjustment factor we used is from the site-specific soil vapor survey in month 3. The survey found the thermal gradient at depth was consistent enough to apply a fixed correction rather than a variable one. That’s why the factor looks nonstandard — it’s calibrated to this site.”
Cully wrote that down too.
Under “Ellison team technical response.”
Grant said: “Exactly right.”
He said it after she finished.
As if he were confirming her addition.
She looked at the map on the screen.
She said nothing else.
The EPA enforcement action notification arrived by email on a Tuesday in March.
Camille read it at 8:30 AM.
Subject: BLACKWOOD SITE — EPA REGION 4 ENFORCEMENT ACTION — COMPLIANCE HEARING SCHEDULED.
The body of the email explained: trace PCB concentrations had been detected in the South Fork Creek water sampling station, 1.8 kilometers downstream from the site boundary.
The EPA was opening a compliance hearing to determine whether the remediation had been conducted correctly.
The email said: “The enforcement hearing will require the submission of all original field sampling logs including chain-of-custody documentation signed by the Licensed Site Assessor of record, in accordance with 40 CFR Part 136.”
She read “Licensed Site Assessor of record.”
She knew what 40 CFR Part 136 said.
She knew it required that supplemental data be submitted by the LSA of record through the EPA Region 4 digital portal, authenticated by the LSA’s personal EPA system credentials.
She had those credentials.
She was the LSA of record.
She had signed 147 chain-of-custody forms.
Grant did not have an LSA license.
He had not disclosed that to the EPA in any of the project documentation she was aware of.
She did not call him.
She opened the project file on her computer.
She scrolled through the chain-of-custody forms.
147 forms.
LSA #GA-4471 on every one.
She closed the file.
She looked at the Dewar flask on the corner of her desk.
The yellow evidence tape on the neck.
The adhesive was lifting at one edge — she had not removed it since the last site visit.
She put the flask in her field case.
She zipped the case.
She set it on the floor by her desk.
She did not call Grant.
She waited.
Grant read the enforcement action email at 9:15 AM.
He read “chain-of-custody documentation signed by the Licensed Site Assessor of record.”
He was not immediately concerned.
He had submitted data to EPA Region 4 nine times over four years.
He had a project manager account on the EPA digital submission portal.
He opened the portal.
He entered his Halcyon project manager credentials.
He selected “Supplemental Data Submission.”
He selected “Blackwood Site.”
The portal displayed a message.
It said: “This submission category requires LSA-of-record authentication. Project Manager access is insufficient for chain-of-custody data submission. Please contact your Licensed Site Assessor to complete this submission.”
He read it.
He read it again.
He called the EPA helpline.
The helpline operator explained, calmly, that under 40 CFR Part 136, chain-of-custody data at a Superfund site could only be submitted by the Licensed Site Assessor of record — the individual whose license number appeared on the original collection forms.
Grant said: “The project manager can’t submit on the LSA’s behalf?”
The operator said: “Not for chain-of-custody data. The regulation requires the LSA of record to authenticate the submission personally.”
He said: “I see.”
He hung up.
He opened the project file.
He opened the first chain-of-custody form.
It was dated March 14, 2021.
It was signed: C. Osei, LSA #GA-4471.
He opened the second.
C. Osei, LSA #GA-4471.
He opened the twenty-third.
C. Osei, LSA #GA-4471.
He had that number on 147 forms.
He had known that number for four years.
He had never thought about what it meant that it was hers and not his.
He had thought about the budget.
He had thought about the schedule.
He had thought about the client relationship and the quarterly reports and the cover letter he had written for the EPA submission.
He had thought about all of those things as the project.
The forms were part of the project.
The number on the forms was on the forms.
He had not thought about what it would mean if the number was the only thing the EPA accepted.
He called Camille.
She picked up on the second ring.
She picked up on the second ring.
He said: “I need you to submit the supplemental data packet to EPA Region 4. Under your credentials. Today.”
She said: “When is the compliance hearing?”
He said: “Three weeks. The submission deadline is Friday.”
She said: “I’ll handle it.”
He said: “Thank you.”
He hung up.
He had not said: I couldn’t log in.
He had not said: the portal rejected my credentials.
He had not said: I read the forms and saw your number on every one of them.
He had said: “I need you to submit the data. Under your credentials. Today.”
She had heard the sentence.
She had understood what was behind it.
She had said: “I’ll handle it.”
She picked up the field case.
She drove to the EPA Region 4 regional office.
The EPA Region 4 office was in Atlanta.
She had been there four times in four years.
Each time for a technical meeting.
Each time she had sat in the row behind Grant.
She sat at a terminal in the data submission room.
She entered her EPA system credentials.
The portal recognized her LSA number: GA-4471.
It displayed: “Dr. Camille Osei, Licensed Site Assessor, Georgia. LSA of Record: Blackwood Site, Halcyon Environmental Group.”
She was named in the system.
She had been named in the system since the project registration in 2021.
She selected “Supplemental Data Submission.”
She selected “Blackwood Site.”
She uploaded the 147 chain-of-custody forms.
She uploaded the full sampling protocol — all four iterations, each with revision dates and her initials.
She uploaded the groundwater flow model.
She uploaded the Henry’s Law temperature correction documentation from the month-3 site report.
She uploaded the quarterly monitoring data summaries.
She submitted the packet.
The confirmation screen read: “Submission received. Submitted by: Dr. Camille Osei, LSA #GA-4471, LSA of Record. Submission acknowledged by EPA Region 4 Enforcement Division.”
She printed the confirmation.
She drove back to the office.
She put the confirmation in the project file.
She did not send it to Grant.
He called her at 4:30 PM.
He said: “The submission cleared. Good. That should satisfy them for the hearing. We’ll address the attribution issue after the hearing settles down.”
She said: “The attribution is already in the submission. My name is on the record as LSA of record and protocol author.”
He said: “Right. Well, that was always the plan.”
She said: “Okay.”
She hung up.
The compliance hearing was a Thursday in April.
EPA Enforcement Attorney Miranda Cole ran the hearing.
She was 44 and had been with the EPA enforcement division for 16 years.
She had read the data submission.
She had read the 147 chain-of-custody forms.
She had read the four protocol iterations.
She had read the groundwater flow model.
She had read the Henry’s Law correction documentation.
She had seen, in the submission metadata, that the submitting LSA was Dr. Camille Osei.
She had also looked up the original Blackwood Method submission on the Superfund database.
She had seen the author listed as G. Ellison.
She had noted the discrepancy.
She had not asked about it in the hearing.
She had asked technical questions about the PCB plume modeling.
She had directed all of those questions to Camille.
Grant was at the table.
He did not speak during the technical portion.
He answered one question about the project budget.
At the end of the hearing, Cole had looked at Camille directly.
She had said: “Dr. Osei. Your protocol and your field documentation appear consistent with the remediation requirements. We’ll have questions as we review the downstream detection data. I’ll send them to you directly.”
She had sent the first batch of questions that evening.
Subject: BLACKWOOD SITE — FOLLOW-UP TECHNICAL QUESTIONS FOR DR. C. OSEI, LSA OF RECORD.
The questions were addressed to Camille.
The cc line included Grant.
He was not in the To line.
Marcus had found out about the hearing from Camille the following week.
He said: “Did he at least say thank you when he called?”
She said: “No.”
Marcus looked at his sample labels.
He said: “Okay.”
He said nothing else.
He had been in the field with her for four years.
He had signed the witness line on 147 chain-of-custody forms.
He had signed his name next to hers.
He knew whose name was the one that mattered.
He went back to the labels.
Before she drove to the EPA office, she spent 20 minutes pulling the project file for the forms she would need.
She had the 147 chain-of-custody forms in the project archive.
She had the four protocol iterations, each time-stamped with the revision date and her initials in the document header.
She had the groundwater flow model, which she had run in a GIS environment on her workstation and exported as a series of quarterly snapshots.
She had the Henry’s Law temperature correction document from month 3.
She had the quarterly monitoring summaries.
She assembled them into a submission folder.
She labeled the folder: BLACKWOOD SUPPLEMENTAL DATA — DR. C. OSEI, LSA #GA-4471 — EPA ENFORCEMENT SUBMISSION.
She had labeled this way on every project submission she had ever made.
It was a habit from her first site assessment, 14 years ago.
Her supervisor then had said: “Put your name and your number on everything. You are the credential. The data is only valid because of what’s on the bottom of the form.”
She had done it since.
She had done it on every form at the Blackwood site.
She had done it 147 times.
She had also, before driving, reviewed what Grant was allowed to have submitted without her knowledge.
The answer was: the executive summary, the cover letters, the client deliverables.
He was the project manager.
Project managers signed the management documents.
She was the LSA.
The LSA signed the scientific documents.
She had understood this division when she joined the Blackwood project.
She had understood that the management documents were his to sign.
She had not thought, four years ago, about what would happen if the management documents also claimed credit for the scientific documents.
She had been thinking about the sampling sequence.
She had been thinking about the groundwater model.
She had been thinking about the PCB plume.
She had typed “Thanks” in the email and gone back to work.
She drove to the EPA office.
She sat at the terminal.
She entered GA-4471.
She submitted.
She drove back.
The project amendment arrived by email on a Monday morning.
Subject: BLACKWOOD SITE — CONTRACT AMENDMENT — PRINCIPAL INVESTIGATOR DESIGNATION.
It was from Grant.
It said: “Per project record update, Dr. Camille Osei is designated as Principal Investigator on the Blackwood Site remediation project, effective retroactively from project initiation in 2021. All future site work under the Halcyon-Blackwood contract will be executed under her PI designation.”
He had added no explanation.
He had not called before sending it.
She read it.
She filed it.
Before the amendment, he had submitted a formal authorship correction to the EPA Superfund database.
She had not known about the correction until the EPA Region 4 system sent her an automated notification.
The notification said: “EPA Superfund Remediation Database — Protocol Attribution Update. The following project has been submitted for attribution correction: Blackwood Site, South Carolina. Corrected attribution: Remediation Protocol developed by Dr. Camille Osei, LSA #GA-4471, Lead Site Assessor. Original submission author: G. Ellison (error). Correction filed by: G. Ellison, Halcyon Environmental Group.”
She had read “error.”
She had read “Correction filed by: G. Ellison.”
He had filed the correction himself.
He had not told her he was doing it.
He had done it before the contract amendment.
He had done it first.
The correction would take 90 days to post to the public database.
She had looked up the update schedule.
It was on a quarterly cycle.
The correction would post in 67 days from the filing date.
She had put the date in her calendar.
In 67 days, the Blackwood Method would be attributed to her on the federal EPA database.
Until then, it was attributed to Grant.
That was the current state of the public record.
She was aware of it.
She had checked.
She had put it in the calendar.
She had gone back to work.
The new contract was a chromium site in Alabama.
She had the project file on her desk.
She had the groundwater model open on her secondary monitor.
The chromium plume at the Alabama site pulled east-southeast at 0.9 meters per day.
She had built the model from the site’s soil conductivity survey, which she had reviewed for three weeks before accepting the contract.
The sampling sequence would begin at the northwest corner — upflow from the plume — and proceed in the direction of movement.
She had written the sequence on a notepad.
She had checked it against the conductivity data twice.
She would field-test the sequence on the first site visit.
The first site visit was in two weeks.
She had the Dewar flask in the field case on the floor next to her desk.
She had added fresh tape to the neck — orange this time, for the Alabama protocol, layered over the yellow from the Blackwood site.
Two colors.
She would pull it from the case at the tailgate on the first day.
She would set it on the tailgate.
She would do the first vapor draw.
The flask would be cold from the ice pack.
She had done this 147 times.
She would do it again.
Grant had also, after the hearing, had two more conversations with the EPA that he told her about by email.
The first: Miranda Cole had called him to ask about the contract structure between Halcyon and the Blackwood client.
He had told Cole the contract was being amended to name Dr. Osei as Principal Investigator.
Cole had said: “For the purposes of the enforcement record, what was her designation during the remediation period?”
He had said: “Lead Site Assessor and protocol designer.”
He had not said “Field Sampling Lead.”
He had said “protocol designer.”
He had said it because Cole was a federal attorney and he had been in a compliance hearing that had just demonstrated, in 90 minutes, that the protocol was Camille’s.
He had said it because the correct answer was available to him and he had given it.
He had emailed Camille about the call.
He had summarized the exchange.
He had written: “I described you as Lead Site Assessor and protocol designer. I think that’s the accurate description.”
She had read “I think.”
She had not replied.
She had gone to the Alabama site schedule.
She had the chromium model to finish.
The second conversation: Cully, the EPA project officer, had reached out to her directly.
Subject: BLACKWOOD SITE — FUTURE MONITORING COORDINATION.
It was addressed to Camille.
Not to Grant.
It said: “For future quarterly monitoring under the post-remediation monitoring plan, EPA Region 4 will coordinate directly with the LSA of record. Please advise on your availability for the Year 5 Q1 site visit.”
She had replied with her availability.
She had cc’d Grant.
He was not in the To line.
She had noted the change in the cc structure.
She had not commented on it.
She had set her availability in the calendar.
She had opened the Alabama model.
She had gone back to work.
The Alabama site was in a light industrial corridor outside Decatur.
The former operator had been a chromium plating facility.
The chromium contamination was in a different chemical class from the PCBs at Blackwood — hexavalent chromium, water-soluble, moving faster in the groundwater than PCB-laden oil.
The remediation approach would be different.
The sampling protocol would be different.
She had spent three weeks building the approach from the site-specific data.
She arrived at 7:30 AM on a Tuesday.
Marcus was already there.
He had the drill equipment out of the truck.
He said: “Which direction do we start?”
She said: “Northwest corner. The plume is tracking east-southeast. We start upflow, then follow the movement.”
He set the equipment.
She opened the field case in the bed of her truck.
She pulled out the Dewar flask.
She set it on the tailgate.
The flask was vacuum-insulated, 500mL, the same flask she had used at the Blackwood site for four years.
The neck had two layers of evidence tape now.
The original yellow layer from the Blackwood final site visit was still there, slightly compressed under the fresh orange tape she had applied this morning for the Alabama protocol.
At the edge where the orange hadn’t quite covered the yellow, both colors showed — a thin stripe of yellow beneath the orange, visible if you knew to look for it.
Under both layers, the LSA adhesive tag carried her number.
GA-4471.
The same number on 147 chain-of-custody forms at the Blackwood site.
The same number that Miranda Cole had used to address her follow-up emails.
The same number that was, at this moment, being processed in the EPA Superfund database correction queue, waiting for the quarterly update cycle that would post in 57 more days.
She did not think about the 57 days.
She was looking at the northwest corner of the site.
She had her sample collection sheet.
She checked the first location against the groundwater flow model on her tablet.
The model was from the site survey she had completed last month.
The chromium plume was moving at 0.9 meters per day in the east-southeast direction.
She had mapped seven sampling locations in sequence: upflow first, then the plume body in the direction of movement, then three downstream confirmation draws.
She wrote the sequence numbers on the flask label in permanent marker.
She wrote: ALB-01-VAPOR.
She checked the time.
7:52 AM.
She said: “Starting ALB-01 vapor draw.”
Marcus logged the time.
She positioned the draw tube.
She pulled the sample.
She capped the flask.
She initialed the chain-of-custody form.
She signed the bottom: C. Osei, LSA #GA-4471.
Marcus signed as witness.
The flask was cold in her hand from the ice pack.
She began the first vapor draw of the new protocol.
The chromium site was different from Blackwood in every technical respect.
The contaminant was different — hexavalent chromium, Cr(VI), a water-soluble metal ion that moved faster in the groundwater than the PCB-laden oil at Blackwood.
The remediation approach was different — she would be recommending a permeable reactive barrier rather than soil vapor extraction.
The sampling protocol was different — the sequence was structured around hydraulic conductivity rather than vapor pressure.
The flask was the same.
She had used it for 14 years.
She had used it at the Blackwood site for four of those years, 147 times.
She would use it here.
She had used it at six sites before Blackwood.
She had used it at two sites in parallel with Blackwood.
The flask had been through perhaps 400 sample collections across those 14 years.
The glass was the same glass.
The vacuum insulation was intact — she had checked it in the field each time by monitoring the sample temperature over a 90-minute hold period.
If the temperature drifted more than 0.5 degrees, the insulation was compromised.
It had never drifted more than 0.3.
The LSA tag had been replaced once, when the adhesive had given out in year 8.
She had put a new tag on.
She had written GA-4471 on the new tag.
The same number she had received in 2015 when she passed the LSA examination.
The number was registered with the Georgia Environmental Protection Division and with EPA Region 4.
The number was on 147 forms at the Blackwood site.
The number would be on the forms here.
The number was on the forms at every site she had worked.
Marcus said: “Are we starting at location 2 next, or do we go to the backup well first?”
She looked at the groundwater model on her tablet.
The hydraulic gradient was running east at this hour, consistent with the morning survey data.
She said: “Location 2. The backup well is downgradient. We take it last.”
He moved the equipment.
She capped the flask.
She labeled it: ALB-02-VAPOR.
She signed the chain-of-custody form.
C. Osei, LSA #GA-4471.
Marcus signed as witness.
She put the flask back in the ice cooler.
She checked the time.
8:19 AM.
She had five more collection points before the midday break.
She had the full protocol to run.
She went to location 2.
She had the flask in her hand.
The flask was cold.
