Calvin opened the spreadsheet and turned the pages toward him. "Find the line for the year-two refurbishment variance," he said. Frank located it after a moment. "You projected fifty-five thousand dollars."
Calvin nodded. "And the current variance?" "Sixty," Frank replied. They sat in silence for several seconds, both aware that the difference between forecast and reality was small enough to support confidence but large enough to require attention.
"We're inside the model's range," Calvin said. "That doesn't make the expense pleasant. It means we planned for the possibility." Frank studied the printed date at the bottom of the page.
"You wrote this in November 2010?" "The first version. I've updated it since, but I keep every older version." Frank turned another page and asked what would happen if three large failures occurred together during the harvest window instead of separately in summer.
Calvin showed him the contingency section, which required the cooperative to retain enough cash for emergency outside capacity and to reduce nonessential spending before touching the repair reserve. Frank read it carefully and looked less angry afterward. "So when I come in here with bad news, you already have a column for it."
Calvin gave a tired smile. "I have a column for a range of bad news. I don't have a column that can promise no surprises. That's why I want you to bring me the work orders before the problems become somebody else's emergency."
Frank moved to the longer-term comparison. Year seven, under the middle service-cost scenario, showed a favorable net position of approximately $2.1 million against the new-fleet purchase. The result included the refurbishment expenses, financing differences, service projections, and lost-time assumptions that Calvin had discussed with the board.
"That's what you're betting on," Frank said. "That's what the evidence says is possible," Calvin corrected. "If the assumptions change, so does the figure. The number isn't a prophecy; it's a way to recognize whether the decision keeps earning its place."
Frank left with the repair authorization and a copy of the current model. Later that week he asked Nora and Ben to develop a list of systems likely to produce similar expensive failures over the next eighteen months. Instead of waiting for each machine to announce its problem, they could inspect during planned downtime and replace parts before secondary damage made the work much more expensive. The result was not a perfect preventive program, but the first useful version of one tailored to the actual fleet instead of a standard schedule printed for machines in average service. Calvin recognized it as the kind of improvement that could outlast his own tenure.
The two board members who had opposed cancellation continued asking questions, and Calvin respected the questions even when their tone made him weary. Edgar Bell pointed out that some of the refurbishment invoices were higher than the projected cost of routine scheduled maintenance on new equipment. Calvin agreed, then asked him to include the new fleet's purchase debt and projected service dependency in the same comparison. Edgar replied that the higher future costs might never occur. Calvin answered that he had always considered that possibility and invited him to examine the optimistic case, which was still available in the original model without any numbers erased.