Dr. Nanduri returned for a second interview and walked Beck through the technical implications. She drew a simple cross-section on a whiteboard: the old rail yard sat on layered fill, and the groundwater table rose during prolonged rain. The approved plan directed water toward a treatment basin with a lined base and monitored outlet. The revised trench shortened the route and reduced construction costs, but it also brought the flow closer to the marsh. If the trench overflowed, sediment and dissolved compounds could move toward a roadside drain. She emphasized that the design change was not automatically unlawful. A different route could be approved if supported by calculations, sampling, and permits. The problem was that the supporting work did not appear to have been completed.
Beck asked about the B-17 station. Nanduri said it was placed near the warehouse because the site had a history of industrial use. A groundwater sample from that location was important to establish a baseline before construction. The field team was supposed to collect a duplicate sample and send both containers to the laboratory. If the results differed, the technician would document the discrepancy and arrange a resample. A missing container should have triggered an incident report. Without a reliable sample, the contractor could not claim that the site was stable based on that station alone.
The investigators compared Nanduri's explanation with the lab's raw instrument files. The B-17 report had been generated using a data set that appeared to be copied from another station. The laboratory's analyst said he had been told that a field technician had delivered the wrong label and that the sample could be matched from the project notes. He had accepted the instruction because the numbers looked plausible and because the project was behind schedule. He now understood that the chain of custody had been compromised. He agreed to provide a sworn statement describing who had requested the substitution.
Rafael received only a limited update. Patel told him that the design review had revealed material inconsistencies and that an engineer had confirmed her name was used without approval. She did not share the full technical report, which remained part of the investigative file. Rafael asked if the findings could help prove that the project had threatened the neighborhood. Patel said they could help establish the need for further sampling and determine whether the contractor's certifications were reliable. The case still required testing of the site and records connecting each person to the decisions.
At home, Isabel asked why an engineering drawing could be more dangerous than a broken pipe. Rafael explained that a pipe could be seen and repaired, while a drawing guided where people built, dug, and moved material. A wrong assumption on paper could become a physical problem after a storm. She asked whether someone could tell from a drawing that it was false. He said sometimes the revision history revealed the change, and sometimes a field measurement showed that the plan did not match the ground. The answer seemed to satisfy her. She returned to the robotics table and began adding a note to her project report about the limits of remote sensor readings. The family had developed its own shared vocabulary for uncertainty, and it helped them keep moving.
The photographs from the original survey team had been stored in a forgotten cloud folder under a project code that no longer appeared in Meridian's active directory. One image showed standing water beside the warehouse access road. A thin iridescent film lay near a concrete channel, though the photographer's note described it as “surface glare.” Another image showed a torn geotextile barrier and dark soil piled against a temporary berm. The metadata placed the photographs on dates when Meridian had certified the site as dry and stable. A third image, taken after a storm, showed a shallow stream crossing the service road and disappearing into vegetation near the marsh.
Beck brought the photographs to an independent environmental consultant, who reviewed them without being told the investigators' preferred theory. The consultant said the visible conditions raised questions about containment and drainage. The film might have been oil, an organic sheen, or a reflection from the camera angle. The dark soil could have been ordinary fill or material from the former rail yard. A torn barrier was a concern because it reduced control over runoff, but a photograph could not establish whether contaminants had moved. The consultant recommended a sampling plan that covered the channel, downstream sediment, and areas near the temporary trench.