'Estimates of convenience': researcher says Alberta spills data too positive

A researcher says the agency that monitors Alberta's energy industry has underestimated the impact of tens of thousands of spills going back decades.

Kevin Timoney, an Edmonton-area consulting biologist, used sophisticated statistical analysis, an extensive research review and comparisons with other jurisdictions to conclude the Alberta Energy Regulator doesn't have a good handle on how much oil and saline water has been released into the environment or remains there.

"Their spill volumes and recovery volumes are too good to be true," said Timoney, who did the study on behalf of northern indigenous bands concerned about spills on their land.

Timoney began with an AER database of 23,655 oil spills and 14,833 primary spills of saline water reported between 1975 and February 2013. The database included records of how much oil or water was spilled and recovered.

Timoney found that the regulator considered 100 per cent of the oil was recovered in 53 per cent of the oil spills. The median recovery rate for oil was 100 per cent and for saline water 80 per cent.

"Thousands of spills reporting essentially perfect oil recovery raise questions of data validity," he writes in a report which was to be released Thursday.

'It's a major liability'

The regulator listed two types of spill environments: air-land or muskeg-stagnant water. Timoney found that recovery efficiency was reported to be the same for both. Nor was it significantly affected by the size of the spill.

Timoney turned to previous academic spills research. In eight major studies five on land and three in water the median recovery rate was 43 per cent. No one documented a perfect recovery.

He then looked at other jurisdictions.

North Dakota, one of the few places where data was available, said only 3.4 per cent of its spill recoveries were 100 per cent successful.

• Pipeline security expert warns of crime risk as new projects proceed
• RachelNotleymulls Trans Mountain pipeline pitch in face of B.C. anger

A graph of U.S. oil-spill volumes revealed a smooth, even curve while Alberta's graph looks like a staircase. That suggests, Timoney writes, "a large proportion of Alberta spill volumes are estimates of convenience rather than measured volumes."

Spill database incomplete

Tracie Moore, a spokeswoman for the regulator, said the database was never meant to provide a complete picture of spill cleanup. She pointed out that before 2014, the regulator shared cleanup responsibilities with others such as Alberta Environment.

"The remaining volume from the release may have been cleaned up through other means and under the jurisdiction of other entities," she wrote in an email.

If a tax auditor looked at this, there'd be people knocking on doors the next day-KevinTimoney, biologist

Timoney also subjected the spill and recovery data to a Benford analysis, a statistical tool that exposes anomalies in large data sets. It has been used in everything from forensic accounting to biology and is accepted as evidence in some U.S. courts.

The AER recovery data was off for both oil and saline water, Timoney found. Given the size of the data set, Timoney concluded the chances of the AER numbers representing actual measured values were vanishingly small.

"If a tax auditor looked at this, there'd be people knocking on doors the next day."

Moore said the regulator relies on industry to accurately report volumes released and recovered.

"This information may then be verified by AER experts," she wrote.

Timoney suggests numbers in the AER's database are the result of managerial decisions and not in-field reporting. That could mean spill and recovery volumes have been underestimated for decades, he said.

"We don't know how much is spilled and we don't know how much is left in the environment."

Regulator confident of improved accuracy

The AER database says that habitat was damaged in less than one per cent of all spills. In contrast, a study of oil and saline releases in Oklahoma from 1993 to 2003 found damage to surface water, crops or livestock, soil, fish or wildlife in about one-third of 17,000 cases.

Again, Moore noted, monitoring habitat damage may have fallen out of the pre-2014 regulator's purview.

"Now that AER has jurisdiction over the entire life cycle of oil, gas, coal and oilsands, we have a better understanding of the impact of incidents on the environment."

• Pipeline leaks due to human error an increasing problem, says NEB
• Enlisting Alberta First Nations to bolster pipeline monitoring

Lingering impacts from spills are common on the Alberta landscape, Timoney said, even on sites that are considered remediated. Effects can persist in residual soil contamination and non-natural plant communities.

He believes his investigation raises major concerns in how the regulator monitors an industry with a footprint in Alberta of 12,000 square kilometres.

"We have an environmental risk that is pervasive across the province and it's unassessed," Timoney said.

"It's a major liability not only for the present, but if we ever try to fix things, for the future as well."