32 INSIGHTS 2018
ALGORITHMS
IMPROVE
RESPONSES TO
ENVIRONMENTAL
INCIDENTS
Robin Ormerod, Managing Director, Envirosuite
When an environmental incident occurs, acting quickly and effectively is the aim of every regulatory
agency and industry involved. Locating the source of the problem and alerting those affected must be
done as swiftly as possible so any impact can be kept to a minimum. It might be a nuisance problem
like a bad odour from a process upset or a more serious incident that could affect health and safety.
I
n many cases, however, achieving the goal of speed and
effectiveness can be difficult. For example, where there
are multiple facilities in an area, complaints about odours
can be aimed at a variety of different locations. Each has
to be investigated and evaluated before the agency is able to
pinpoint what is causing the problem and where it is located.
In the meantime, staff may have been dispatched to the field to
deal with the event without a clear idea of the true situation.
This challenge is exacerbated when weather conditions —
particularly wind — change quickly, a situation that happens
often in many places. The wind direction occurring while the
investigations are being carried out could well be different from
that which occurred when the incident first happened. This can
make tracking down the source more difficult and time consuming.
An example of this challenge occurred in the San Francisco
Bay Area in December 2016. During that event, more than 1400
complaints were received from residents about strong petro-
chemical odours. The odours resulted in some 120 hospital
admissions of people suffering from headaches, burning eyes
and sore throats.
Investigations into the source of the odours took months
before it was finally traced to oil that had leaked during the
unloading of a tanker kilometres away. This news was small
comfort for all those who had suffered at the time but could
not at that stage be told exactly what had happened.
Taking an algorithmic approach
Clearly a better approach is needed to manage environmental
incidents, and this can be achieved through appropriate software.
By making use of sophisticated algorithms based on the best
science, agencies can radically transform the way they respond
to such events. They can also help companies that experience
such incidents respond in a more proactive manner, reducing
the chance of such events in the first place and improving the
speed and effectiveness of response. This reduces the negative
impact on nearby citizens and can help to build trust around
longer term concerns that might be present in the community.
When an incident occurs, one quick and powerful step enabled
by such software is to generate a real-time trace of where the
odour or pollution has travelled from. The algorithms take the
time and location of received reports and combine them with data
about current and recent wind and other atmospheric conditions.
In this way, the likely path of the offending release can be traced
back to the source much more quickly than has previously been
possible — within seconds rather than days or weeks.
This approach can also be used by industries that experience
an event that results in odours or pollutants being released
off-site. The tools also can generate a forward trajectory map
showing where plumes are likely to travel and the areas that
will be most affected, allowing more effective management of
response activities.
Significant operational benefits
For environmental regulatory agencies, adopting such analytical
and predictive systems can deliver significant benefits. When
an incident occurs and complaints are received, the source of
the problem can be found much more rapidly. The tools also
reduce the need to send officers out into the field without some
prior information, which can waste considerable time. Instead,
a back trajectory can quickly determine the likely source and
resources sent there to investigate immediately. Time saved in
fruitless searches for offending sources can be considerable.
