Over the next couple of days we began to look at his system. Historical data that was more than five years old had been archived, but had at least been grouped into folders labelled land, air, surface water, effluent and waste. Opening these folders revealed a large number of spread sheets and databases, which were effectively gathering dust. Whilst there was lost value in that data from the viewpoint of historical trends and so on, that was not the main issue for the V.P.
Many types of current data was still being collected; groundwater (levels and quality), waste production, dust and other air quality variables from the emission stacks (VOCs, NOx, SOx, particulates), surface water, effluent quality and flow rates, and weather. Data was also being collected on water use and recycling, fuel use (all sources) and waste being recycled and landfilled.
Some of this data was manually collected, some resulted from analysis by external labs and some was automatically collected from data loggers. Because the growth in data management had happened organically, different, and disparate, processes had evolved in parallel to deal with this. The processes were in part driven by the data source (e.g. the format in which data was provided by a lab or data logger) and in part by of the idiosyncrasies of the people involved.
The net result was that lots of people were doing lots of things but working inefficiently. Regular reporting of data was error-prone and reports were often late. The reporting process was matter of cutting and pasting, sometimes from five different spread sheets, to make a single report.
Our conclusion was that the V.P. was right – his data was in a mess and, without intervention, was likely to worsen. One such operational practice is usually an on-going monitoring program, with defined targets for many of the measured variables. The frequency of the monitoring can be anything from continuous (e.g. for dust, flow rates, vibration or weather), hourly, daily, weekly, monthly, quarterly, half yearly or annually. Breaches need to be addressed and reported to the control authority, sometimes within 24 hours of the incident.
Remedial measures need to be put in place, and repeated exceedances can, in extreme cases, result in prosecution, fines or even withdrawal of permits or shutting down a facility. The data also needs to be reported internally and externally in differing formats and at regular intervals (monthly, quarterly, biannually or annually).