Data is not the full measure

Michael Scott of SWIG describes steps being taken to improve the reliability of operator self-monitoring in order to establish regulator and public confidence in measurement data

• measurement selection,
• sampling considerations,
• installation design,
• installation, calibration and maintenance.

In the water and wastewater treatment and environmental industries data is sparse, both spatially and temporally. Two reports have shown that confidence is not only low in measurement data but, because of a lack of standards and agreement on how to express it, there little or no comparability between member states in the EU:

• pre-normative research for ISO TC147 and an EU report by Anders Lynggaard Jensen concluded that comparability of data between member states is not possible. This was one of the inputs to ISO/TC 147/WG2 for ISO/CD 15839 Water quality - On-line Sensors/Analysing Equipment for Water Specifications and Performance Tests of which the author has seen version N29 and suspects this is now out of date. The number of versions seems to demonstrate the difficulty of defining specifications and performance tests for the wide range of measurements needed in the water and wastewater treatment industry and also the application sensitivity of their use,
• a 1999 UKWIR report, On-line Instrumentation Standards & Practices, concluded there is a wide spread lack of confidence in on-line measurements. Measurement is multi-discipline and multi-cultural because of the need to
  consider:
• measurement device/methodology design - the rapidity of technology change,
• business support - the speed of business change,
• application design,
• data transmission,
• data use,
• data storage,
• standards/best practice support.

In addition, Glen Harvey, chief executive officer of the Instrumentation, Systems, and Automation Society (ISA) has just reported on a year-long USA industrial tour. He concludes that process control data and control techniques are more important than ever, but instrument engineering disciplines have all but disappeared. Many observers consider the picture is similar in the UK and
the rest of Europe.

Regulation needs
The monitoring of process effluents and wastewater discharges and their impact on receiving environments is a key element of regulation under environmental protection legislation, and of the provision of information to the public on pollution control.

Historically, the majority of effluent and wastewater monitoring was undertaken by the appropriate regulatory authorities. In the years following the introduction of the Environmental Protection Act 1990, industrial process operators have been required by conditions in environmental permits to carry out compliance monitoring programmes and report the results to the Environment Agency (EA). This has become known as operator self-monitoring. Particular emphasis has been placed on the use of continuous monitoring systems. This approach is being extended through implementation of the Pollution Prevention and Control Regulations and the Urban Waste Water Treatment Directive.

In order to ensure this approach commands the confidence of both the regulator and the public, the EA is pursuing several initiatives to help improve the quality and reliability of operator self-monitoring. The self-monitoring of flow initiative started with the issue of P150 Flow Monitoring of Discharges: An audit Manual.

As part of the development of the monitoring certification scheme (MCERTS) the EA is in the process of extending the scheme to continuous water monitoring systems (CWMS). These monitors are often used for the monitoring of discharges of process effluents and wastewaters (for example from industrial processes and sewage treatment works) and of receiving waters (for example rivers, lakes and reservoirs). The MCERTS's performance standards and conformity testing requirements were published for an eight-week consultation period on January 7, 2002.

The EA's key drivers for the scheme are to:

• monitor wastewater discharges and receiving water quality in order to assess the environmental impact of the industries it regulates,
• ensure the EA's performance requirements for water monitors are clearly stated,
• improve the reliability of monitoring results.

It is believed the MCERTS will ensure instrument manufacturers will demonstrate the manufacturing process is controlled under a quality management system to produce instruments that deliver consistent performance. Both the EA and the industrial process will then have mutual confidence and the surveillance costs will be reduced.

Sampling
Data used for establishing that regulations and consents have been met is almost all derived from samples taken from the point of measurement and subsequently analysed in a laboratory. It is no surprise that automatic samplers are already regulated under the EA's E32 document against which a number of samplers have been approved and are in use. There are a number of detail questions yet be resolved on the transition from the E32 approval to MCERTS accreditation. Even with samplers there are application issues. Tests are done on clean water and yet some samplers will be used on dirty water. Dirty water must be defined.The point at which the sample is taken must be demonstrably representative. Some commentators point out that each application has quirks and that reliability and maintenance problems will take a number of years to appear. One of the major suppliers is convinced the only way forward is for the supplier to monitor the sampler remotely and take care of application and maintenance issues as they arise.

Investment
Monitoring and process measurement is certainly less than 1% of capital investment so senior management can be forgiven for not considering it to be a strategic issue. However self-monitoring will mean investment in measurement will become an important senior management consideration. The water industry must, for Ofwat reasons, make plans for at least five years, but realistically needs to take 10 and 20 year decisions on its asset investments. Careful selection and use of process and monitoring equipment and support services can have a disproportionately beneficial or detrimental affect on capital and revenue budgets. The footprint of a new plant and existing plant capacities and/or output qualities can be increased and hence civil engineering costs significantly reduced with the application of new monitoring and control technologies. The costs of demonstrating compliance with regulations and directives could be reduced and the risks of adverse publicity also reduced with a broader look at the monitoring assets and the support costs required.

Business would benefit if:

• the net present value of each measurement was compared with the business value of that measurement,
• a very strong case was made for measurements for which the above is not favourable,
• the quality of measurement data were an integral part of the use of the data.

The problem with the introduction of new approval/accreditation demands is there are inevitably going to be significant costs and this may well inhibit the introduction of new products and technologies. There are many important small suppliers in this market sector and the accreditation costs might be prohibitive in a market noted for small margins.

work to be done
The EA and the trade association for instrumentation, control, automation and laboratory technology in the UK - GAMBICA, has made a good start with draft standards and an accreditation facility for monitoring, but further work needs to be done on the people-skills part of the uncertainty equation. While data quality will certainly benefit from the EA's MCERTS initiative much more work needs to done on design and application practices and on the all-important business analysis of support costs and comparison with the business benefits.

The recently reported work on auditing flow metering systems carried out by the national engineering laboratory (NEL) on behalf of six UK water companies is also encouraging but not extensive enough.

The water and wastewater treatment monitoring market is not large and is highly fragmented by application, with a significant number of suppliers. All these needs and concerns are becoming apparent at a time when it is becoming extremely difficult for companies to spare the appropriately skilled and experienced managers to work on the consensus specification and best practice document generation.

It is suggested that to help deal with both the above challenges it would make good business sense to organise some pre-competitive co-operation to generate:

• updated/new standards,
• accreditation of instruments, design practices and people,
• best practice guides.

The views expressed in this article are those of the author and must not be taken as policy statements by, in particular, SWIG or the EA