Monitoring in World Bank Projects
Monitoring of environmental quality is often included in World Bank projects, to help understand the state of the ambient environment or to monitor the emissions and impacts of specific discharges. Monitoring is usually only a small component (except in TA projects) but it is often important in measuring and evaluating the outcomes of a project. Feedback on the success of these components has been limited to date but this is changing as more emphasis is placed on monitoring the impacts of projects. However, comments in a number of World Bank reports indicate that information available on the environment is often incomplete or unreliable (see Box). While technical consultants can advise on the design of systems, ensuring the long term effectiveness and performance of such systems is much more difficult.
Insights from a Sample of Bank Reports
The systems are often not providing useful data:
'Data and information systems (physical, technical, socioeconomic, etc.) relating to water resources in terms of quantity, quality, accessibility and use are generally inadequate throughout the region' African Water Resources
'With adequate information, setting priorities is not difficult, requiring only a comparison of benefits and costs. But environmental data are generally incomplete, so uncertainty about costs and benefits is high.' East Asia’s Environment
'In most parts of the country there is no basis upon which to make informed decisions about ambient environmental conditions.... Filling the gaps is a precondition for an assessment of pollution costs in these areas.' Argentina Pollution Study
A number of causes have been identified:
'... reduction of budgetary support; lack of understanding of the economic importance of the data; and TA programs which are unsustainable because the outputs have not been of use to decision-makers and therefore programs are not funded.' Sub-Saharan Africa Hydrological Assessment
'High turnover rates [of high quality trained technicians] have been a major problem ... and have contributed to the intermittent operation of a sophisticated network of ... monitoring stations build with [donor] financing.' Thailand Country Report
'The old system of central planning and control has left a legacy of inefficiency and mismanagement resulting from: unreliable basic information; narrow segregation of responsibility; poor information dissemination and analysis; incomplete accountability for performance and results; and in certain cases, deliberate misreporting of environmental data.' Anon.
The most important factor in establishing or upgrading an environmental monitoring system is to have agreement on the objectives of the system and to design the system to address these objectives. A monitoring system should be designed to provide practical management or scientific information, and better information will improve environmental decision making (up to a point). At the same time, collection of data, maintenance of a database and carrying out appropriate analyses is costly (in terms both of human and financial resources). It is therefore important to focus resources and priorities on those areas where the information is most needed and most useful.
Monitoring is always included in the preparation and design of major projects which may have a direct environmental impact, such as power stations and sewage treatment plants and there are other examples from the portfolio where specific support has been given to monitoring components. In these more focused cases, success with monitoring has typically been higher.
Examples from the Portfolio
Shanghai Environment Project: Water quality monitoring was a significant part of this project, where a key objective was the relocation of the water supply system intake to a point in the river where industrial pollution was a minimum threat to the supply. In addition, a sophisticated satellite and GIS system was provided to track urbanization in the catchment area.
Lake Victoria Environmental Management Programme: One major objective of this project, covering the three major lake countries of Kenya, Uganda and Tanzania was to collect and share the data necessary to understand the dynamics of the lake system.
Bolivia Environment, Industry and Mining Project: An example of a cooperative approach where detailed sampling and analysis of a mining area were carried out with bilateral (Swedish) assistance, separately from the Bank project but as part of the same overall government program.
Brazil National Industrial Pollution Control Project: This project focused on improving industrial pollution control in the city of San Paolo and included a component to strengthen the data management capabilities of the local agency.
Ambient Data and Emissions Data
The conceptual model which underlies most pollution management that emissions of pollutants lead to changes in ambient levels, which in turn control the impacts on health and environment. The ultimate concern is the impacts but in practice, ambient data are often used to provide information on background conditions and as a basis for policy setting. For the design and control of a specific project it is usual to work with emissions data because this is more easily measured and managed at a specific site. The emissions requirements, however, must be related to estimates of the overall impact on the ambient levels (and ultimately on the environment).
The links between emissions, ambient levels and impacts need to be very carefully understood when a monitoring system is being designed because an error in the assumed relationships can lead to wasteful or counter-productive policies and actions.
Ambient monitoring is carried out for a variety of reasons including assessment of environmental problems and evaluation of interventions. The initial design of a program is usually based on available (often unreliable) data on existing conditions or sometimes on simple models based on emissions estimates. In any case, the program should have flexibility to be adjusted in the light of initial results.
The choice of parameters should be based on the sources in the area and on the receptors and impacts of concern. In practice, it is usually worthwhile measuring a basic set of parameters (see Box) plus any others of special concern. The monitoring plan should set out the rationale for selecting the number and location of monitoring stations, the monitoring frequency and the sampling methods, as well as a quality control plan. The design of monitoring systems should not be over-ambitious: even in the most advanced countries such as the USA and EU, the management information available from large scale monitoring systems is less than would be hoped. Such experience reinforces the benefits of beginning with a small, focused monitoring system, concentrating on answering key management questions.
A realistic set of monitoring parameters would normally include the following (although the exact requirements will vary with specific circumstances.
Basic set: Suspended particulate matter (preferably including PM10 or PM2.5): SOx; NOx: Lead.
Others: Ozone: VOC: Aerosol Acid.
Basic set: pH; DO; BOD; Suspended solids; Flow (if appropriate).
Others: Coliforms; Ammonia; Nitrogen; Phosphorus; Chlorophyll; Nitrate; Metals.
Environmental Monitoring Singapore.
In order to provide an overview of the environment in this island state, Singapore regularly monitors six key air pollutants (PM10, SOx, NOx, O3, CO and hydrocarbons) at about 15 main sites and monitors three major water quality parameters (DO, BOD and SS) at about 70 locations around the island.
Emissions monitoring is usually carried out for the design and operation of pollution control systems or for regulatory purposes. For operational purposes a small number of parameters (or surrogates) may be measured on a regular or continuous basis, while regulatory requirements are typically very specific as to a sampling schedule.
Emissions monitoring should include measurement of flow rates although a surrogate (such as production rate) is often used. Flow measurements are necessary to convert measurements of concentrations into estimates of pollutant loads. Continuous monitoring methods are now available for many of the most important air and water pollutants but the value of additional data obtained needs to be weighed against the cost and complexity of such systems.
Environmental quality assessment: This type of assessment is essentially a baseline study, either for the examination of the impacts of a project (in a formal Environmental Assessment) or as a basis for the preparation or examination of policy options. In the more sophisticated type of assessment, cause and effect relationships are estimated so that the impacts of different interventions can be determined. In such an assessment, very large amounts of useful data and analyses are often obtained but the details are frequently then stored in a form or location where subsequent access is not easy
Monitoring as a System
Monitoring usually refers to tracking trends over time and must be considered as a system comprising a number of elements, where the overall quality of the system is controlled by the weakest segment.
Sampling refers to the collection of data which is representative of a system. In some cases the data can be measured directly (such as temperature) but often the representative sample has to be analyzed or tested to determine the value of individual parameters. Important questions are the design of the sampling scheme and the protocols for the collation, storage and transportation of samples. A wide range of national, international or sector specific standards exist for sampling and analysis (e.g. series such as ISO, EN, BS, DIN, API, etc.).
Analysis of samples is a critical step and the value of the monitoring results depends to a considerable extent on the level of confidence which can be put on the analysis. In many cases, a major issue is the capability and credibility of the laboratory system which will be used for the analysis.
Information management refers to processing the data obtained from the sampling system. This includes recording the data, analyzing it and presenting the information in a manner which is useful to decision makers and other stakeholders.
Monitoring the Vistula River in Poland
The Vistula River has been monitored since the 1970s, with the results being used to classify the state of the river. The basic monitoring program involved 35 stations on the main river and over 500 monitoring stations on the tributaries. The samples (a standard set) were analyzed in 50 local laboratories across the country. Given the large amounts of data being collected, there were concerns over the quality of the results and a new set of 5 key permanent monitoring stations has been set up on the Vistula, together with a certified laboratory testing program, to provide a highly reliable set of baseline data.
The choice of sampling methods should always be made on an evaluation of factors such as reliability, accuracy ease of operation and cost.
Documentation from the GEMS/Air program provides an indication of the tradeoffs to be made between simple (often labor intensive) methods and more sophisticated approaches.
The keyword for analytical systems is simplicity. The developing world is littered with sophisticated laboratories, funded by donors or projects, which are either not operated because of a lack of funds for simple items (such as glassware or purging gases) or are highly unreliable -- often because the laboratory buildings cannot be kept at constant temperature or dust- and contaminant- free. The problems are commonly compounded by the lack of a national standards infrastructure which could grade or certify the laboratories.
Experience has shown that an incremental approach is often the best way (see Box), where the capabilities and reliability of existing laboratories are gradually strengthened and expanded. In this approach, a major emphasis is placed on maximizing the use and productivity of existing systems and in implementing Quality Control systems, before introducing new equipment or capabilities. External quality control, from either national or international bodies, is critical to establishing the credibility and competitiveness of individual laboratories.
Many World Bank projects have included a component to finance laboratory equipment for pollution monitoring and training of personnel. One which has been successful is the Poland Environment Management Project. In this case, a Polish speaking external expert, with many years experience in managing laboratory systems, was brought in to inventory existing facilities and to optimize the use of existing equipment. A Quality Control system was also introduced before decisions were made on the expansion of the laboratories and the purchase of new equipment. The laboratories were encouraged to operate as far as possible on a commercial basis and to broaden their client base beyond the state agencies that had traditionally been served. The project also included support for national standardization efforts, designed to increase the reliability of the overall laboratory system in the country.
Definition and Collection of Data
Decisions on the data required and its collection will be influenced by a range of factors including the existing data (and its quality); local capabilities in sampling and analysis; the existing information infrastructure (such as the ready availability of remote sensing data); the projected life of the monitoring system; and, of course, the costs of establishing and maintaining the data collection system. The costs of collecting and entering data can be many times the costs of the hardware and initial training.
Data Handling Systems and Information Management
The determination of institutional responsibilities for the handling and management of information is frequently a difficult issue. A pragmatic approach is that the organization which needs the data should initially collect it (or contract for it to be collected); that the initial processing should be as simple and straightforward as possible (for example using spreadsheets or simple database software on a standard PC); and that the data should be stored in a format which is simple and convenient for exchange, once agreement has been reached.
'Building an Environmental Information System requires an new institutional culture' World Bank paper on SSA
More elaborate systems (often based on GIS systems, for example) need to be founded on institutional agreements about technical issues (such as the geo-referencing system) and about exchange and interpretation of data. The installation of a number of GIS systems in different agencies or organizations is not necessarily inefficient but care must be taken to avoid duplication and to ensure compatibility.
The costs of a monitoring system include both capital costs and operating and maintenance costs. The capital costs of equipment can be estimated reasonably reliably but operating costs are often highly dependent on local labor costs and the difficulties of obtaining spare parts.
A 1993 USEPA estimate of air pollution monitoring costs indicated an annualized figure of around $26,000 for continuous monitoring of some key pollutants (See Table) but these costs are generally coming down as equipment is improved.
Estimates for the establishment of pollution control laboratories in India (for a 1991 World Bank project) were $0.22m for a regional laboratory, $0.14m for a mobile laboratory, also $0.14m for a continuous ambient air monitoring station and $0.011m for a continuous water monitoring station.
Estimates for a 1993 project in the Ukraine included $2.2m for 16 stationary air quality monitoring stations (about $0.14m each); $1.3m for 7 mobile ambient air quality monitoring stations (about $0.19m each); and $1.9m for 7 mobile emissions monitoring vans (about $0.27m each). In addition, sample costs for measurement of toxics deposition were estimated to be of the order of $200-500 each sample for PCBs, mercury and PAHs.
A basic ambient air quality monitoring program for a large metropolitan area is based on 6 automatic monitoring stations around the city, each measuring sulfur dioxide, carbon monoxide, ozone, nitrogen oxides and non-methane hydrocarbons, while another 16 manual monitoring stations measure particulates (as PM10) and sulfur dioxide. A composite air quality index is prepared and announced daily, with contingency plans implemented when the index is very high. The performance of the system is audited twice a year by an internationally reputable laboratory. The annual cost of the system is estimated to be less than $1m.
Given the costs in time, and in human as well as financial resources, it is essential to establish responsibility for the collection of data and maintenance of the information systems. Collection of the data is almost always the most costly component of a monitoring system and it is unrealistic for an environmental agency or a national statistics office to attempt to collect large amounts of data. Collection should be the responsibility of the line agencies responsible for various functions such as water supply or transportation so that in this way the operating costs of the monitoring group can be minimized. The coverage of data may be less than desirable or optimal but the system is far more likely to be at a level which is sustainable in the long term.
In any case, it is essential that the environmental monitoring unit has an assured budget to sustain the effort -- reliance on donor funds for set-up is acceptable but the ongoing operations must be funded to a realistic level by the country. Examples of the practical budget problems that have been encountered include lack of fuel to drive project vehicles to sampling sites, inability to pay for long distance phone calls to regional centers, and lack of operating budget to pay for basic consumables such as glassware and distilled water. Difficulties can arise where governments seek grant money for the development of new monitoring systems because such systems then tend to be capital intensive and over-sophisticated and therefore unsustainable in the long run. However, with more awareness of these concerns, there is increasing emphasis in project design on working within realistic institutional and budgetary constraints.
Monitoring is usually though of as a somewhat complex technical issue, to be carried out by experts. However, there is increasing interest in developing simple systems which can be adopted by communities to monitor their own local environment. These systems can be based on simple technologies (the equivalent of a strip of litmus paper) or can be more
sophisticated and involve the training of local technicians in basic sampling and testing procedures. In either case, the involvement of the community in the design and implementation of the system is critical. Experience with these approaches is gradually developing and is likely to expand in Bank projects with greater community or NGO involvement.
'What Is A Minimum Data System?'
The fundamental question which should be and often is asked is 'what is a minimum data system?' for a given situation. There is not a simple answer but a number of basic principles can be set out to assist in the design of an environmental information system.