Lakeland Mills appeared to be making progress on the challenges it was facing after adding a third shift to its sawmill to process more beetle-killed pine before a dust-related explosion leveled the facility nearly three years ago, a coroner’s inquest heard Thursday.

Alan Little and Glenn Roche died from injuries suffered in the April 23, 2012 disaster and more than 20 others were injured, some seriously.

Bill Barwise, who was the sawmill’s operation superintendent at the time, told the inquest that midway through 2011, 57 more employees were hired for the new shift.

Prior to that point, Lakeland ran two shifts of 10 hours each from Monday to Thursday. The third shift added up to an additional 36 hours over Friday, Saturday and Sunday.

Barwise said beetle-killed pine would break apart more easily than the spruce and balsam it also ran, creating more debris and dust. By November 2011, “the problem was at its worst.”

“We were running in excess of 90 per cent pine and had issues with the baghouse,” Barwise said.

But once a new sorter had been installed, Barwise said they were able to mix in green wood, whenever the dust had become too great for the mill’s dust collection system to handle.

The system consisted of hoods placed over top of the mill’s machinery that would suck up dust and send the material to the baghouse through ventilation pipes.

He said it would take about 15 minutes to switch to a different species – workers were give dust masks for the interim – and once completed, it would take about 10 minutes for the dust to settle down.

“It alleviated a little bit of the dust so to me, it felt like we were moving in the right direction,” Barwise said.

There were fewer downtime hours available to clean and maintain the mill. In answer, Barwise said such steps as adding workers to cleanup detail and adjusting shifts, particularly over the weekend, were taken.

Barwise was later questioned about a major fire on Jan. 19, 2012 at the large headrig where Roche had been working.

The headrig was one of two used to break down logs. The inquest has heard a saw “deviated” sending up sparks that lit dust on the machine’s overhead hood then sending a fireball as high as 50 feet into the air.

Roche and small headrig operator Brian Primrose put out the flames with their extinguishers while some fellow employees rained fire retardant down on the two from above, the inquest has heard.

Barwise confirmed that Roche told him “we almost lost the mill.”

“And I told him I never want to see anyone get hurt in the mill and rather than ever seeing you put yourself in peril, I’d rather see the mill go up than for you to be hurt extinguishing it,” Barwise said.

The incident was not reported to Prince George Fire Rescue because there was no structural damage and no one was injured but Lakeland three supervisors wrote a report and handed it to Barwise three days later.

However, the report indicates none of the workers who had saw the incident were interviewed although Barwise said one of the supervisors was there at the time. The report said the dust collection system was not operating properly and there was inadequate cleanup in the area.

Barwise said he followed up by making sure there was extra cleanup on the graveyard shift to blow down the headrig. Barwise said he also gave Little, a supervisor, permission to shut the mill down and clean up the debris if there was too much around.

“He had my full support,” Barwise said.

That occurred twice under Little’s watch over the next week.

The next day a blast similar to the one that struck Lakeland three months later hit Babine Forest Products, also killing two people and injuring more than 20 others. Barwise said people at Lakeland were very concerned about what happened at Babine but did not know what to do.

He said dust and debris were considered fire hazards but it was only after the Lakeland disaster that he learned how explosive it could be.

“How do you manage when you don’t know?,” Barwise said. “You don’t understand what the root of the issue is.”

In earlier testimony, Garth Turner, the sawmill’s maintenance superintendent at the time, testified problem spots in the mill were “remarkably clean” on the morning of the fatal explosion.

Turner said that at about mid-morning he had checked the spots he specifically asked be cleaned up over the weekend in the debarker and slasher areas.

“I was impressed with the way it was and that was after the mill had been running for 12 hours,” Turner said.

Turner’s workday ended at 6:30 p.m. on April 23, 2015, which was a Monday, and slightly more than three hours later, a “deflagration” or subsonic explosion erupted, destroying the sawmill.

His comments drew an expression of disbelief from one Lakeland employee in the gallery and coroner’s counsel John Orr later pressed Turner on the statement, saying his perception was so different from what other witnesses have testified so far.

Turner responded that in contrast to many other sawmills, Lakeland had a dust collection system – the material was sucked up through hoods placed over top of the mill’s machines and sent to a baghouse – and those without one could be “very dusty.”

“The desire would be to have a perfectly clean mill but the reality is it didn’t happen, it couldn’t happen,” Turner said. “When the suction system was working, it was a very clean mill.”

Turner said he did not recall seeing much airborne dust when he left for the night.

Asked if a decision to increase the number of shifts from two to three affected the performance of the bag house, Turner said the volume of intake per hour did not increase, only the number of hours it was in use.

And he said misting systems, used to remove dust from the air around a new chipper that had been installed as well as the mill’s debarker worked very well.

He noted the mill featured beams made out of laminated lumber that had no top edge for dust to rest on. However, he also conceded they could be difficult to clean because they were hard to get to.

He also denied ordering “flash cleanups” or “impress the inspector” cleanups. Prince George Fire Rescue fire prevention officer Steve Feeney testified he noted a marked improvement in the mill’s condition when he returned in March for a look following an inspection in November 2011 where the dust and debris was so bad he took photos.

Turner said he believes other factors contributed to the blast, namely methane, but admitted he had not read the WorkSafe B.C. report which goes into some depth on the steps taken to test for the gas and why it was found to not have played a role.

Both Turner and Barwise no longer work at Lakeland.