Risk & Innovation: Small Companies in Six Industries: Background Papers Prepared for the NAE Risk and Innovation Study (1996)

Charles E.Bangert Jr. and Richard A.Lynch

The environmental lab industry is unique in that the chief driver of the industry—governmental regulation—also acts in some ways to inhibit innovation and slow the pace of technological advancement in dealing with environmental problems. There is little question that the comprehensive set of environmental laws that have been passed over the past 20 years has forced Americans to address the issue of environmental degradation with more vigor than if purely market forces had been at work. At the same time, however, the lack of flexibility inherent in governmental regulatory schemes has acted in many cases to remove incentives for environmental testing labs and instrumentation suppliers to explore innovative approaches to improving monitoring capabilities for environmental contaminants.

This lack of strong incentives to “push the envelope” in adopting new technologies within the industry has limited the opportunities for small, entrepreneurial companies to force paradigm shifts in technology. Consequently, instrumentation manufacturers in recent years have focused mainly on incremental product improvements that enhance the efficiency and lower the costs of providing

analytical services. A few relatively large companies dominate the industry by providing these incremental improvements at the lowest cost possible.

The laboratory business itself, on the other hand, has always been fragmented. This is somewhat curious because there are demonstrable economies of scale inherent in operating a laboratory. Because of the high fixed costs associated with establishing and operating a lab, it is advantageous to have high sample throughput to spread the fixed costs over the largest number of samples possible. Available evidence suggests, however, that smaller laboratories have been successful because administrative, quality assurance/quality control (QA/QC), customer service, and data management problems increase as laboratories grow; after a certain point, these problems begin to outweigh the advantages of scale in production efficiency.

There is reason to believe that the status quo regarding environmental analytical technologies is changing. The U.S. Environmental Protection Agency (EPA) has been active in recent years—through programs such as the Superfund Innovative Technology Evaluation (SITE) program—in attempting to promote a more flexible approach to the application of new technologies to environmental assessments and cleanups. In addition, the agency is slowly moving toward allowing the use of performance-based testing methods (PBMs) to replace some of the rigid, highly prescriptive testing methods currently in use. Widespread adoption of PBMs (which is still probably years away) would remove some of the barriers to innovation that currently exist by providing labs and instrument manufacturers with some of the incentives that are currently lacking. Finally, market conditions are forcing environmental labs to become more aggressive in using technology to reduce costs and boost operating efficiencies. As a result of all of these factors, the pace of technological innovation in environmental testing may soon begin to accelerate, with some significant new technologies emerging from smaller, entrepreneurial companies.

The tremendous advances being made in computers and computer software are also profoundly affecting the laboratory industry. The “product” of environmental laboratories is, after all, information. Samples of water, air, or soil are the raw materials that are processed to provide clients with information regarding the nature and extent of environmental contamination. Until recently, however, the environmental lab business and instrumentation suppliers have been relatively slow in harnessing the available advances in computational power. Individual lab instruments have possessed computer control and data collection capabilities for some time, but the next step of integrating these instruments into laboratory-wide information networks is only now beginning.

The purpose of this background paper is to: provide an overview of the environmental testing laboratory industry; discuss likely future trends in testing technologies and methods; and provide a framework for discussing the implications of these technological trends for the industry.

The roots of the environmental testing industry can be traced to 1962, when drinking water standards advanced by the U.S. Public Health Service spurred the development of water-quality testing labs. This nascent industry consisted of small in-house water utility labs as well as commercial “mom and pop” laboratories that tested for a variety of simple parameters such as water hardness, metals, and coliform bacteria. The methodology and instrumentation used for these analyses were fairly simple.

Industry growth accelerated with the introduction of the first two major pieces of environmental legislation that called for extensive analytical testing: the Federal Water Pollution Control Act amendments of 1972 (now referred to as the Clean Water Act, or CWA) and the Safe Drinking Water Act of 1974 (SDWA). (The testing requirements mandated by various environmental statutes will be discussed below.) These two pieces of legislation expanded the amount of water and wastewater testing being performed, increased the complexity of analytical procedures, and required more sophisticated instrumentation. Because much of the testing required under the CWA and the SDWA was accomplished by in-house municipal laboratories, however, the commercial environmental laboratory industry remained relatively small until the late 1970s.

A second wave of major environmental legislation increased the demand for commercial environmental laboratory services exponentially, starting in the late 1970s and early 1980s. The two seminal pieces of legislation were the Resource Conservation and Recovery Act of 1976 (RCRA), which regulates the treatment, storage, disposal, and transportation of both hazardous and nonhazardous solid wastes, and the Comprehensive Emergency Response, Compensation and Liability Act of 1980 (CERCLA, also known as Superfund), which requires the cleanup of abandoned hazardous waste sites. With the introduction of RCRA and CERCLA, and subsequent reauthorizations and amendments to all major environmental laws, the commercial laboratory industry enjoyed rapidly rising revenues and impressive profitability virtually throughout the entire decade of the 1980s. Technological innovations were of the incremental variety, involving lower detection limits for contaminants and the application of microprocessors to improve ease of use and data-generating capabilities of instruments.

Most observers of the environmental testing industry believe that the industry began to plateau around 1988. Growth rates averaging 25 percent or more during the mid-1980s have slowed to around 5 percent currently. A number of factors have contributed to this slowdown, some of which are probably transient. The sluggish economy of recent years, for example, has resulted in many industrial companies cutting back on nonessential environmental compliance activities. This trend has been exacerbated by a reduction in regulatory compliance and enforcement activities by state and federal agencies, primarily because of budgetary problems. These factors are inherently cyclical and should decline in importance at some point.

On a more fundamental level, however, the industry has changed permanently. Testing regimens under existing environmental legislation have matured (albeit with several notable exceptions) and are averaging single-digit (or low double-digit) percentage unit growth. Most of the major environmental laws are due for reauthorization over the next 2 to 3 years, but there is little chance that sweeping new testing mandates will be included. Meanwhile, two areas that were expected to generate substantial new business opportunities—the Clean Air Act Amendments of 1990 (CAA) and the cleanup of hazardous waste and mixed-waste sites on federal facilities owned by the Department of Defense (DOD) and the Department of Energy (DOE)—have developed slowly.

Declining unit growth rates, in combination with an excess of analytical testing capacity, have steadily eroded prices and profitability in the laboratory industry over the past several years. As a result, technological advancements in instrumentation have become increasingly focused on incremental improvements in operating efficiency and ease of use. Greater emphasis is now being placed on using computers to improve productivity through automation of lab procedures and through better integration of the various laboratory instrument systems.

The current size of the commercial market for environmental testing services is probably between $1.5 billion and $1.6 billion, with annual revenue growth at less than 5 percent. (Note that revenue growth is lower than unit growth because the average price per test continues to decline.) Testing under RCRA accounts for about half of total industry revenues, Superfund-related testing another 20 percent, CWA 15 percent, and SDWA 5 percent, with the balance being divided among various other environmental statutes.

Reliable data on the size of the market for environmental instrumentation are difficult to obtain. The same analytical instruments that are used in environmental testing are also used for other purposes such as academic or pharmaceutical research, and instrument suppliers do not usually provide sales data broken down by end user. Wade Miller Associates estimate, based on the best available data, that the current size of the U.S. environmental instrument market is $500 million. This includes some environmental instrument sales for nonlaboratory purposes such as on-line water quality monitoring. Sales of instruments to labs probably accounts for only half of total sales, or about $250 million annually.

Reliable estimates of the growth rate of instrument purchases are likewise difficult to obtain, but it appears that instrument sales have generally tracked with environmental labs’ revenue growth rates. In the mid-1980s, when the lab industry was growing rapidly, instrument sales were also increasing 15 to 20 percent annually. In recent years, as industry growth has leveled off, instrument sales have increased at a much slower rate, about 5 percent or less per year. Anecdotal evidence suggests that

instrument sales actually might have declined slightly in 1992 and 1993, as labs reduced capital spending in the face of persistently weak market conditions.

The total number of commercial analytical testing laboratories in the United States is between 1,400 and 1,600. As stated above, the industry is very fragmented. Only 25 to 30 environmental testing labs generate revenues of greater than $10 million per year. These large labs together account for approximately 30 percent of the industry’s total revenues. An additional 80 to 100 labs generate annual revenues of between $3 million and $10 million. These labs, which generally have a regional (as opposed to the larger labs’ national) orientation, collectively account for about another 25 percent of industry revenues. The remaining 45 percent of revenues are accounted for by 1,300 to 1,400 small, local labs with revenues of less than $3 million each. The largest environmental laboratory company in the United States is Enseco Inc. (a subsidiary of Corning Inc.) with approximately $90 million in annual revenues. Only two other laboratories—IT Analytical and National Environmental Testing (NET)— have more than $45 million in annual revenues. (Both of these companies are also subsidiaries of larger corporations.)

On May 2, 1994, Corning and International Technology Corporation announced plans to combine their environmental laboratory operations. If the deal is consummated (it is subject to regulatory approval), the joint venture will be by far the largest environmental laboratory in the country, with approximately $150 million in annual revenues. As will be discussed in greater detail below, this proposed merger may be the first in a series among the largest lab companies.

The analytical instrument industry is much less fragmented than the laboratory industry, and it is also very much a global industry. According to the Environmental Business Journal (August 1992) the top 25 manufacturers of analytical instrumentation (including nonlab instruments) account for about 75 percent of the total global market; the top three alone account for almost 30 percent of the market. The top three environmental instrument manufacturers are Hewlett-Packard, Perkin-Elmer, and Thermo Instrument Systems (Finnigan), each of which generates (by EBJ’s estimates) more than $175 million from environmental instrument sales worldwide. About half of the environmental instrument sales of the “big three” are in the United States, and all three are part of larger companies with significant resources. Smaller instrumentation manufacturers, of which there are probably several dozen in the United States, tend to specialize in niche markets (e.g., Isco in wastewater samplers) or emerging technologies (e.g., Ensys, HNU Systems, and Spectrace Instruments in field testing technologies).

Consulting engineering companies are the most common purchasers of environmental analytical services, accounting for more than 40 percent of industry revenues. Consulting engineers work principally to remediate environmental problems of industrial companies, although engineers also conduct assessments and cleanup work for government clients. Direct analytical work for industrial clients accounts for another 30 percent of revenues, and work for federal, state, and local governments accounts for most of the balance. Of the federal government work, only about 10 percent is contracted through EPA, with most of the rest deriving from cleanup at DOE and DOD installations.

RCRA provides guidelines for the storage, transport, and disposal of solid waste from both municipal and industrial sources. Both hazardous (Subtitle C) and nonhazardous (Subtitle D) wastes are regulated under RCRA. The HSWA greatly expanded the programs established under Subtitle C to include thousands of smaller hazardous waste generators, expanded “corrective action” regulations (concerned with remediation of RCRA sites), created Subtitle I to regulate underground storage tanks, and mandated a reevaluation of Subtitle D requirements for managing nonhazardous landfills.

CERCLA was enacted to clean up abandoned hazardous waste sites and hazardous spills that pose a danger to human health and property. SARA significantly increased the scope and financial resources of the program. CERCLA and SARA are collectively known as Superfund, under which was established a trust fund to provide money to clean up sites for which a potentially responsible party cannot be identified. The sites that pose the most urgent threat to human health and the environment are placed on the National Priorities List (NPL) and are the responsibility of the EPA Superfund program. All other sites, referred to as non-NPL, are the responsibility of the Superfund programs of the various states. Abandoned hazardous waste sites on federal property are handled by the agencies involved.

The intent of the CWA (i.e., the Federal Water Pollution Control Act amendments) was to restore and maintain the quality of the nation’s water. The primary mechanism to achieve these goals is the National Pollutant Discharge Elimination System (NPDES) permit program. Under this program,

any point source of pollution discharge into any U.S. waters—either directly, or through a public sewage plant—must apply for and receive a permit. These permits require the point source to disclose the volume and nature of its discharges, establishes concentration limits on the discharges, and requires monitoring and compliance reporting. The 1987 WQA amended the CWA by increasing control over emissions of toxic chemicals and required development of sludge management and combined sewer-overflow permitting programs.

SDWA is the legislative framework for the establishment of national drinking water regulations. Under SDWA, EPA sets minimum standards for drinking water regulatory programs that are administered by the states. The centerpiece of the drinking water standards is maximum contaminant levels (MCLs), which are established by EPA for a variety of contaminants. Water systems are required to perform regular monitoring of water quality to ensure compliance with the MCLs, and to report publicly any violations. The 1986 amendments to the SDWA tightened the requirements by: replacing the previous discretionary authority of the EPA to promulgate MCLs with a mandate to do so for 83 specific contaminants; requiring filtration of all surface water supplies and disinfection of all public water systems; and toughening the enforcement posture of EPA

The CAA of 1970 (and subsequent amendments in 1977) aimed to improve the quality of the nation’s air by establishing National Ambient Air Quality Standards (NAAQS) for reducing ambient levels of ozone and other pollutants and providing for the implementation of national standards at the state level through the development of State Implementation Plans (SIPs). SIPs divide each state into Air Quality Control Regions (AQCRs), each of which is required to have a compliance strategy for meeting the air quality standards.

The 1990 amendments to the CAA contained 11 separate titles that significantly broadened air quality regulations. The four titles that will have the greatest impact on demand for analytical testing services are: Title I (Nonattainment), which increases enforcement activities for AQCRs that are out of compliance with air quality standards; Title III (Air Toxics), which greatly expands the scope of the National Emission Standards for Hazardous Air Pollutants (NESHAP) program; Title IV (Acid Rain), which requires coal-fired power plants to reduce emissions of sulfur and nitrogen oxides; and Title V (Permit Programs), which requires states to establish permit programs for air pollution sources similar to the NPDES program for water pollution.

As briefly discussed above, environmental laboratories can be divided into three general classifications based on size and general characteristics: (1) large labs (generally above $10 million in revenues) usually have a national orientation and the ability and inclination to serve large, high-volume customers with sophisticated testing needs. Large labs, for example, conduct a large percentage of the testing related to the Superfund program as well as most of the testing for large national manufacturing corporations; (2) medium-size labs (between $3 million and $10 million in revenues) often pursue niche marketing by tailoring services to regional companies’ testing requirements. The sophistication of these labs is generally on a par with the larger labs, although they tend not to serve the very high-volume customers; and (3) small labs (less than $3 million in revenues) serve a narrower range of local and regional customers and tend to concentrate on simpler testing procedures.

Interestingly, above a certain size, there may actually be diseconomies of scale for environmental testing labs. Medium-size labs appear to be more profitable than either large or small labs. Small labs tend to conduct simpler testing procedures and process lower volumes of samples, so it is no surprise that they tend to be the least profitable. But why do large labs appear to be less profitable than medium-size labs? The exact reasons are unclear, but there are two probable causes: costs, and market and customer mix.

Large laboratories tend to have much higher cost structures than medium-size labs. There appear to be two underlying factors accounting for this disparity in costs: Personnel costs, the largest single component of costs for most labs, are much higher for the large labs than for the medium ones; and administrative and overhead costs, which appear to become a progressively bigger problem as labs grow larger. Part of the reason for the higher administrative burden for the larger labs is that as labs grow and expand into new programs and geographic areas, the costs and complexities of maintaining the proper certifications multiply greatly. (Certification and accreditation will be discussed further below.)

Large lab companies, because of their national, high-volume orientation, more often cater to customers that are technically demanding but, at the same time, these labs are acutely focused on cost as a purchasing determinant. Medium-size lab companies, on the other hand, tend to specialize in regional, niche markets where pricing pressures are not so severe. Customers within these niche markets, which mainly consist of smaller manufacturing companies dealing with wastewater or hazardous waste regulations, tend to focus on customer service issues when deciding which lab to use.

For these customers, services that regional labs are more likely to provide, such as sample pickup and delivery, are at least as important as cost.

Both of these cost problems emphasize the need for better use of computers in laboratories to automate manually intensive procedures, such as QA/QC, sample preparation, and compilation of customer data reports from raw data. These inefficiencies become more apparent as the size and complexity of lab operations grow and are, therefore, more noticeable in the large laboratories. Reducing costs and increasing the flexibility, speed, and responsiveness of analytical testing are perhaps the key immediate challenges facing environmental testing laboratories today.

One other possible explanation for the higher productivity of the medium-size labs is that the large labs may not be large enough to take full advantage of the economies of scale that exist. The largest environmental lab company (Enseco) currently generates less than $100 million in revenues. In relation to other industries where economies of scale have been demonstrated, such as capital goods manufacturing, a $100 million company is relatively small. The recently announced merger of Enseco and IT Analytical will heighten the pressure on many of the other larger, underperforming laboratory companies to either get much larger, through mergers with other large companies, or to downsize significantly. Over the next several years, it is possible that the environmental lab industry will further develop into a two-tier industry, with a small number of relatively large labs generating the bulk of the industry’s revenues and a large number of small, regional labs filling specialty niches.

One of the major issues facing the environmental testing industry is the lack of a national accreditation program for laboratories. Oversight of the performance of laboratories is currently the responsibility of a patchwork of state and federal agencies that impose widely differing—and, at times, inconsistent—certification requirements for laboratories participating in their programs. This arrangement has proven cumbersome, redundant, and very costly, particularly for laboratories that work in several states or program areas. A number of states, for example, require separate certifications for drinking water, wastewater, and hazardous waste analyses. Each certification usually requires the laboratory to run performance evaluation samples, undergo audits, and pay certification fees. This results in significant downtime for the laboratories involved, in addition to the direct costs of becoming certified. Furthermore, because there is no national standard-setting organization for laboratories, the comparability of analytical data from state to state and program to program is questionable.

There is broad agreement among regulators, customers, and laboratories that a national laboratory accreditation program covering all the major federal environmental laws is desirable. Since 1990, a coalition of commercial and government labs, industrial companies, regulators, and others

called the National Environmental Laboratory Accreditation Coalition (NELAC) has been lobbying for the development of a national accreditation system. NELAC has proposed that a desirable national accreditation system have the following characteristics:

• federal oversight of a program administered by third-party accreditation organizations

• uniform, well-defined accreditation standards

• independent review of laboratory data

• on-site assessments of laboratory capabilities

• performance evaluation testing

• provisions for accountability and enforcement

Support for national accreditation is building at EPA. In 1992, an EPA advisory panel, the Committee for National Accreditation of Environmental Laboratories, released a report recommending that EPA develop such a program. A number of hurdles must be overcome before national accreditation becomes a reality, however, including financing issues (i.e., what level of fees the lab industry would be willing to pay to replace the current system) and the role that the states would play in such a program. The laboratory industry is unlikely to support a national accreditation program unless the new system replaces the current hodgepodge of state certifications. Some states are likely to surrender control over lab certification only grudgingly. For these reasons, and because there are currently a number of higher-priority items of concern for EPA, the development of a comprehensive national laboratory accreditation program is unlikely before the end of the decade.

Innovation in environmental laboratory instrumentation has followed four cycles that essentially parallel the four stages of evolution of the industry itself. Technology dictates the types of contaminants that can be analyzed, the concentrations at which they can be detected, and the capacity of instruments to collect and process information. The different “waves” of environmental legislation, therefore, have generally corresponded to different stages in the evolution of instrument technology. It should again be stressed, however, that the application of advanced technology for environmental testing purposes tends to lag the actual development of such technology because of the highly regulated nature of the industry. Technologies have generally been developed in academic or industrial labs for other analytical purposes, then adapted later for specific environmental applications.

The testing instruments used by the early water-quality labs of the 1960s were simple in terms of both the types of measurements being made and in the quality of information generated. Water

quality parameters of interest were mostly aesthetic ones such as color, odor, pH, and hardness, with some testing for the presence of microbial contaminants such as fecal coliform. The most common type of analysis was colorimetric measurement by spectrophotometer, and raw data were provided to the laboratory analyst through the application of analogue meters or strip chart recorders. The transformation of those raw data into useful information then required manual data manipulation and analysis.

During the industry’s growth phase of the late 1960s and early 1970s, the instrumentation used by the industry became progressively more sophisticated. The CWA of 1972 expanded testing to include analysis of wastewater for the purpose of compliance under the NPDES program, and the SDWA of 1974 greatly expanded the range of contaminants for which water utilities were required to test. Some of the more sophisticated techniques that began to be applied during this period were gas chromatography (GC, used for analysis of some types of volatile organics), gas chromatography combined with mass spectroscopy (GC/MS, to test for more complex volatile and semivolatile organics and pesticides), and atomic absorption (AA) or inductively coupled plasma (ICP) spectrophotometry (both used for analysis of metals and other inorganics).

The late 1960s and early 1970s also saw the first widespread use of computers with laboratory instruments. The initial systems were stand-alone minicomputers that could be linked to laboratory instruments to log and process data and, in some cases, to provide rudimentary instrument control. These systems reduced some of the manual tasks associated with data analysis and information generation, but a great deal of “artistry” by talented analysts was still required to generate meaningful analytical data.

The enactment of the RCRA (1976) and Superfund (CERCLA, 1980) statutes catalyzed an enormous increase in the complexity of environmental analyses. Superfund, for example, requires identification and quantification of a broad number of hazardous wastes in a variety of matrices, including soil, solids, groundwater, and air, or some complicated combination of matrices. The biggest market for analytical testing services under RCRA is related to the Subtitle C “cradle-to-grave” regulations governing hazardous wastes. Subtitle C testing includes characterization of hazardous wastes at various stages of generation, transport, storage, and treatment. As with Superfund, RCRA testing requires a high level of analytical expertise to deal with complicated mixtures of wastes in a variety of sample matrices. These two pieces of legislation vastly increased the use of GC/MS as an analytical tool.

Along with increasing the complexity of analysis, RCRA and Superfund (as well as SDWA and CWA, under which increasingly stringent monitoring requirements were being imposed during the 1980s) also resulted in a quantum leap in the volume of environmental analyses being conducted. The surge in volume required a transformation of the industry from a small, research-intensive business into a high-volume production industry, a process which is still taking place. This process has been facilitated by the application of more powerful computers into the lab environment. With the

development of microprocessors in the late 1970s came a new generation of instruments with computer controls that performed automatically many of the diagnostic procedures, instrument control, and information acquisition and manipulation tasks that previously had been accomplished by the instrument operator.

As the laboratory industry has hit a plateau over the past several years, the pace of development of new analytical technologies and techniques has slowed somewhat. However, new technologies and new applications for existing technologies have emerged and will continue to grow in importance. For example, as a greater number of complex, nonvolatile, and polar organic compounds are regulated, the limitations of GC and GC/MS methods for these analytes is requiring the increased

application of technologies such as high pressure liquid chromatography (HPLC), liquid chromatography linked with mass spectroscopy (LC/MS), and sequential mass spectroscopy (MS/MS). The past several years have also seen increased use of autosamplers and other laboratory automation techniques and the expanded use of information management software and other productivity-enhancing tools.

Table 1 presents the applications of analytical instrument technology by environmental laboratories. These technologies, for the most part, do not represent cutting-edge science. Techniques such as GC, GC/MS, and HPLC have been used for many years to conduct routine chemical analyses. To a large extent, this inertia reflects the difficulty of adapting new techniques in a heavily regulated industry that is also prone to litigation. Analyses performed by environmental testing laboratories form the basis of all remedial and compliance activities undertaken by both industry and government. For this reason, the “legal defensibility” of analytical data is of paramount importance. Laboratories, therefore, take great pains to ensure that EPA-approved analytical methods are followed to the letter, which tends to stifle creativity in devising new analytical techniques.

To use one example, the adoption by environmental laboratories of solid-phase extraction (SPE), a technique used to separate organic constituents from aqueous samples, has been slow. This despite the clear advantages that SPE possesses over classical liquid-liquid extraction (LLE). SPE is less labor intensive than LLE, uses much less solvent, and requires less skill on the part of the laboratory analyst. The lethargy of the laboratory industry in adopting this technique can be traced primarily to a lack of “approved” analytical methods for water and wastewater analyses that make use of the technology. SPE is starting to become widely accepted now only because the manufacturers of the devices have worked diligently with regulators over the past several years to gain approval for SPE as an alternative extraction technique.

As has been mentioned, there are signs that legislators and regulators understand this problem and are taking steps to improve the adoption of novel technologies. The previously mentioned SITE program, for example, was established under the Superfund amendments of 1986 to promote the development of new treatment and monitoring technologies. Similar technology advancement programs exist for drinking water, groundwater remediation, DOE cleanups, and other programs. In addition, it is likely that EPA will be more flexible in allowing the use of performance-based testing methods (PBMs) in the future. PBMs (discussed further, below) allow the use of any appropriate technology to achieve a data-quality objective. PBMs should allow laboratories to apply new technologies more flexibly than under the current “standards-based” methods, which rigidly prescribe allowable analytical procedures for particular types of tests.

Regardless of regulatory developments, powerful forces are driving the laboratory industry and its suppliers toward a greater degree of innovation. Foremost among these forces are cost and flexibility. Labs are currently under tremendous competitive pressure to reduce the cost and improve the turnaround time for analytical services. The reasons for this pressure are discussed below, but the important point to be made here is that technologies that can lower costs and improve operating efficiency are those most likely to be quickly adapted. In light of this, the three technology trends that are likely to have the biggest impact on the industry over the next several years are: improved information management systems; increased automation; and increased use of portable field-testing technologies.

The significant expansion of air testing activities that will eventually come about as part of the 1990 amendments to the Clean Air Act will also assist in accelerating the pace of innovation in environmental testing. The CAA has not been a big driver for environmental testing services because the initial legislation passed in 1970 did not result in regulations that required much analytical testing. The amendments passed in 1990, however, should significantly increase the market for air monitoring later in this decade, as the new regulations begin to take effect.

Automation (the use of robotics for sample handling and processing) and improved information management (the use of computer systems to automate data acquisition, analysis, and reporting) are discussed together because they both have been made feasible by recent advances in computer hardware and software technologies. This is probably the most important trend facing laboratories, because it affects virtually every aspect of laboratory operations. Laboratories are no different from many industries today. Competition is fierce, and there is tremendous pressure to reduce costs to stay profitable.

Computers and software have been used with great effectiveness in laboratories to control the operation of instruments, record data, and assist in data analysis and report generation. The major elements of an automated laboratory typically include the following:

• a centralized comprehensive database management system, known in industry parlance as a laboratory information management system (LIMS)

• data systems for collecting and processing raw data from different instrument systems

• automated systems for the preparation and processing of samples

• a network of personal computers that can be used to analyze data, generate reports, and manually enter data

Perhaps the most significant trend in laboratory information systems is the ongoing effort to “network” the elements of automated laboratories into seamless systems. Most labs use instrumentation from a variety of vendors that have traditionally used incompatible, or “closed” operating systems. This lack of compatibility creates tremendous problems when the instruments need to be linked together to share data. The current trend is toward open systems, usually using UNIX operating systems, and the development of instrument interfaces that enable instruments to “talk” with each other as well as with the LIMS. Eventually, the goal is to provide modular instruments that are standardized and interchangeable.

The continued development of more flexible and powerful LIMS software is a second major trend in information systems technology. LIMS systems are the heart of any automated laboratory, and there are a number of LIMS software packages commercially available. These packages tend to suffer from a variety of shortcomings, however. A lack of compatibility with instrument data systems is one such problem, requiring some manual keying of data from the instruments into the LIMS. In addition, many LIMS need substantial customization to be useful for specific applications. For labs that are too small to be able to afford a dedicated systems analyst or programmer, the unavailability of off-the-shelf LIMS programs is a significant handicap. Cumbersome report-generating capabilities, another common problem, often force labs to develop additional capabilities of their own. Finally, most current LIMS have only a one-way flow of data. LIMS systems can generally collect and process results downloaded from instruments, but data that are already in the database cannot usually be accessed to assist in providing real-time benchmarks, for example, or to set instrument calibrations automatically.

Instrument manufacturers are also likely to continue to develop new technologies that automate sample handling and processing tasks that are now accomplished manually. Automation of chemical analyses is already common. Most GC and GC/MS, for example, have automated sample injection and control capabilities. The focus now seems to be turning toward automating sample preparation. The federal government has supported two projects recently that have demonstrated the feasibility of automating these processes: the National Institute of Standards and Technology (NIST) demonstrated a robotic system for the preparation of inorganic samples for spectroscopic analysis; and the Department of Energy’s Contaminant Analysis Automation program demonstrated a similar system for automating the preparation of samples for organic analysis and the cleanup of the resulting waste extracts.

In addition to sponsoring research projects, the EPA is currently developing a series of guidelines for laboratory automation called Good Automated Laboratory Practices (GALPs). The GALPs, which are now being finalized, should stimulate labs to increase the pace of automation by removing some of the uncertainty about what constitutes acceptable automated procedures.

The development of portable testing technologies that can be used at environmental testing sites will have a major impact on the environmental testing industry in coming years. Field testing technologies, to this point, have mainly been used as screening tools to assess whether a predetermined contaminant is present within a set concentration range. The data are essentially used to narrow the number of samples that require testing by more sophisticated fixed-base laboratories. Field testing techniques have been limited to this primary role both because of limitations in the current field testing systems and a reluctance on the part of regulatory authorities to permit these relatively new technologies to be used without confirmation by more established techniques. As field testing technologies continue to improve, however, and as regulatory authorities continue to become more flexible in considering alternative technologies, the advantages in flexibility and cost offered by field testing are likely to result in increasing market share for these systems.

Several different types of field testing technologies are being developed. The following four are most likely to have a major impact on the industry:

• enzyme immunoassays

• portable GC, GC/MS, and other instruments

• mobile labs

• fiber optic (and other) sensors

Enzyme immunoassays seem to have captured the most attention. Immunoassays, which derive from technology originally developed in the biomedical field, take advantage of the specific binding properties of antibodies to identify the presence of contaminants that are specific to a particular antibody. The concentration level of the analyte is determined through a colorimetric procedure that uses a simple field spectrophotometer. The advantages of this technique are that it is much cheaper and faster than conventional techniques, does not require that the analyte be separated from the sample medium (which eliminates sample preparation), and offers better (lower) detection limits than some types of fixed instruments.

Immunoassays have several disadvantages. For example, because the antibody is specific for a contaminant, it is necessary to know what contaminants are present before testing begins. As well, the antibodies developed so far are specific for classes of compounds, rather than specific contaminants. There is also a high incidence of false-positive results. New generations of immunoassays are addressing the last two problems, however, and the EPA is aggressively pursuing approval for wider applications for immunoassays. It is very likely, given the cost advantages of immunoassays over traditional laboratory testing procedures, that immunoassays will continue to grow in importance.

Portable lab instruments and mobile labs are also likely to become more important in the future. Mobile labs consist of temporary, dedicated labs that are erected at a long-term cleanup site and are

removed when the cleanup is finished. Portable instruments, on the other hand, are scaled down versions of the instruments in fixed labs and are able to be transported to the location the testing is required. Both offer advantages over fixed labs in terms of costs and flexibility, with some trade-off in accuracy.

A third type of field testing technology, remote sensors, is still in the experimental stage. These sensors use a variety of technologies, such as fiber optics, for directly measuring concentrations of contaminants in situ. Once developed, sensors could be used for tasks such as continuous monitoring of groundwater for contamination at hazardous waste sites.

The 1990 amendments to the Clean Air Act will increase the demand for laboratory air testing and monitoring services. Title I, for example, expands requirements for “nonattainment” urban areas to monitor for ozone and ozone precursors. Title III vastly expands requirements for industrial sources to monitor toxic emissions.

Unlike many pieces of environmental legislation, the CAA amendments will push technology development, because current air monitoring instrumentation is probably not adequate for generating the quantity and quality of testing data that will be required. It should be noted, however, that the market for air-testing instruments is likely to develop very slowly. The first new air-toxics rules, for instance, are not scheduled to go into effect until 1997, and it is possible that date will slip. Among the areas being explored for the future, however, are the following:

• developing field-based chromatographic instruments capable of unattended sample collection, analysis, data reduction, and data transfer;

• improvements in the current generation of stainless steel air-sample canisters, including using smaller stainless steel or glass tubes filled with adsorbent; and

• development of more efficient continuous emission monitoring equipment (CEM).

PBMs deserve some additional elaboration because of the potential for these types of methods (if adopted) to accelerate the pace of technological innovation within the industry. The current policy of the EPA is to prescribe specific, very detailed “reference” methods for compliance with the monitoring requirements under each of the major environmental programs. In some cases (SDWA, for example), the use of reference methods is absolutely required. For other programs (such as RCRA), the reference methods are intended to provide “guidance” for laboratory analysis. Even when the reference methods are intended for guidance purposes, however, environmental laboratories are loathe to stray even in the tiniest details from the methodology because of fear that the data, if

challenged, will not be legally defensible. The application of a rigid, inflexible testing methodology has resulted in a number of major problems within the industry. These are described below.

For most programs, the EPA is required to include specific testing methods in new regulations as they are promulgated. For these programs, EPA cannot legally change any reference method without going through a lengthy, formal rule-making process, including an official notice in the Federal Register and a public comment period. Delays in implementing new methods can also be attributable to EPA’s tortuous technical review process for new methods developed by commercial labs. As a result of these two factors, there is usually a lag of 5 years or more from when new analytical methods are developed to the point where they are implemented by environmental laboratories. The industry, therefore, is usually using methodology that is one to two generations behind the state of the art.

The intended purpose of environmental analytical testing is to provide data that are useful for decision making. Such data clearly need to be as accurate as possible to ensure that appropriate decisions are being made. In many instances, the analytical methods employed by labs need to be optimized (for differences in sample matrices and other real-life factors) in order to obtain accurate data. Many current methods, however, are written in such excruciating detail that changing the experimental conditions to optimize the accuracy of the results is not possible. The result is data that can be misleading or inaccurate, despite being obtained through an acceptable analytical procedure.

The current rigid methodology employed for a majority of environmental testing restricts the flexibility of labs to modify existing procedures for reducing costs or to substitute less expensive new technologies (such as SPE or immunoassays) for more expensive older technologies. This lack of flexibility has been a major factor in the inability of commercial labs to reduce costs quickly enough to avoid erosion in profit margins as prices for analytical testing have continued to fall (see below).

The solution to these problems, in the eyes of many in the industry, is for EPA to move from its current prescriptive approach toward the increased use of PBMs. PBMs differ from the current methods in that data-quality objectives are more important than the processes used to achieve them. Under a PBM system, EPA would establish performance criteria, such as the desired measurement objectives and quality control requirements, for particular types of tests. Each laboratory would then

have the option of using available reference methods to meet these objectives or developing alternatives or modifications that are better, cheaper, or faster.

PBMs are supported by a majority of commercial environmental labs as a way of improving the health of the industry, the accuracy of the data produced, and (most important for this discussion) speeding the introduction of innovative technologies into the industry. The regulatory community, however, (as well as many customers of analytical services and a minority of laboratories), has been wary about embracing PBMs. Detractors argue that PBMs will reduce the ability of the EPA and customers to evaluate the quality of work of various laboratories and enable laboratories to scrimp on QA/QC procedures. Both of these issues can probably be addressed through the careful design of the performance criteria and through the use of proper QA/QC safeguards, such as a national accreditation program and periodic performance evaluations of laboratories. Because of this resistance, however, the development of PBMs is likely to be slow in the immediate future. The only area where the development of PBMs appears likely over the next several years is for SDWA-related testing. The next rule-making package for SDWA is due at the end of 1994, and it is likely to include at least some PBMs.

The environmental laboratory industry has been in a period of consolidation and intense competition since revenue growth rates began to plateau in 1988–1989. At the industry’s peak in the late 1980s, there were approximately 1,800 commercial environmental labs. There are now between 1,400 and 1,600 labs, and that number is still declining. Laboratories that will survive will be those best able to address the following major challenges.

The primary factor behind the consolidation trend has been a sharp decline in prices for analytical testing services. Prices have been eroding because of an imbalance between the supply of and the demand for laboratory analytical services. In the late 1980s, annual revenue growth for the industry averaged 25 percent or better, and most observers of the industry believed that this rapid rate of growth would continue. Additional testing capacity was added rapidly, both by existing laboratories and by new entrants into the business. However, for a variety of reasons, including the slow development of CAA and federal facilities testing, economic recession, lax regulatory enforcement, and the environmental “policy drift” of a new presidential administration, revenue growth rates slowed sharply in the early 1990s. The double-digit revenue growth of the 1980s has been replaced by growth rates of 5 percent or less. With the industry geared up for continued rapid growth, the result was

inevitable: too many labs chasing too few samples, leading to sharply plunging prices as labs scrambled to generate enough business to cover their investments in new testing capacity.

To students of technology development, the above story is familiar. As technology advances in any field, it tends to create large and rapidly growing markets, which attract new competitors. At a certain point, the technology matures, competition intensifies, prices begin to fall, and a shakeout ensues. The environmental testing industry adds a unique twist because, as discussed above, in addition to technological advances, the industry is strongly affected by governmental regulation. It can be argued, in fact, that the industry probably enjoyed a longer period of prosperity than it otherwise might have because regulatory inertia retarded its natural evolution.

The shakeout in the environmental laboratory business is now occurring, however, and it probably will continue for some time. Prices—which for some types of testing are off by 50 percent or more from peak levels—continue to drop. There are apparently still many labs that, in a desire to keep operating until industry conditions improve, have been practicing low-ball pricing. Revenue growth rates will likely improve from current levels over the balance of the 1990s, but a return to double-digit growth rates is unlikely. Barring a sharp upturn in sample volumes, more lab closures and mergers such as the planned merger of Enseco and IT Analytical, are likely over the next few years.

As prices have declined and revenue growth flattened, profitability also has fallen. Several recent surveys have indicated that average pretax profitability for environmental labs has declined from 10 to 15 percent of revenues to 5 to 7 percent (or possibly lower) over the past few years. The laboratory industry, on the whole, appears to have a cost structure that is too high to maintain adequate profitability in the current pricing environment. Cost control, therefore, promises to be one of the key competitive issues in the industry over the next few years. Firms that can reduce costs and improve efficiency most effectively will be among the survivors of the current shakeout.

To some extent, the high costs are due to factors that are beyond the control of those in the industry. Improvements in instrument capabilities, for example, have led to the development of costly new testing methods for trace-level contaminants. Expenses related to becoming certified in various state- and federal-agency laboratory programs, as well as increasingly stringent QA/QC requirements, have also added to costs.

The single largest component of costs, however, is personnel expenses. Employee salaries, wages, and benefits constitute over half of a typical lab’s expenses, and that percentage has risen sharply in recent years. Short-term remedies for reining in employee-related costs include reducing staffing levels and limiting compensation increases. Those methods are frequently effective in “stopping the bleeding” but do not address the root cause of the industry’s profitability problem: poor productivity. Despite the widespread adoption of computers in the industry, productivity growth has

lagged because lab procedures have not been adjusted to take full advantage of the computational power available. The next great competitive challenge in the industry, therefore, will be to harness the full power of computers to improve lab productivity. Labs will need to take a hard look at eliminating lab procedures that do not add value to the analytical process, at better integrating lab instruments into LIMS, and automating procedures to the maximum extent possible.

The challenge of redesigning operations for the information age is not unique to the laboratory industry. Productivity growth among providers of services in the United States badly lagged that of manufacturers virtually throughout the 1980s despite immense investments in new computers and software. It has only been over the past few years that productivity growth in services has begun to improve as companies take a hard look at their processes to determine which ones are essential to providing effective service and how those essential processes can be made more efficient. The laboratory industry is really a microcosm of that larger trend.

The development and adoption of PBMs—when it eventually occurs—will also assist in this process of redesigning lab operations. The rigidness of EPA-sanctioned testing methodologies currently limits labs to changing only those procedures that are not integral to the actual testing methodologies. PBMs, however, should allow labs much more flexibility in modifying or replacing inefficient testing procedures.

PBMs should also enable labs with talented technical personnel a chance to improve profitability by differentiating themselves from labs with less talented personnel. A critical paradox of the laboratory industry over the past several years is that customers and regulators have demanded ever-increasing quality from providers of analytical testing services (primarily because of the need for the data to be legally defensible), but they have not been willing to pay higher rates to those labs with the most stringent QA/QC procedures. In large part, this is because it is very difficult to judge the relative quality of different labs’ services when all labs are required to conduct analyses according to the same methodology. Customers, therefore, have tended to view the provision of lab services as a commodity item, despite the legal sensitivity of the resulting data. As labs are given more freedom to explore alternative methodology, however, it will be possible for labs to either develop value-added services for which customers will be willing to pay higher rates or reduce costs through streamlining and improving testing procedures.

Companies that provide hardware and software to automate environmental laboratory procedures clearly will be the beneficiaries of this drive to reduce costs. As discussed briefly above, automation of chemical analyses is practiced on a large scale throughout the industry. The next step will be to better automate the handling and preparation of samples and improve integration of the various instruments required for specific testing procedures. A modular approach to instrumentation is

likely to evolve in which all the different types of instrumentation required for a particular analytical method will be linked into highly automated units that will process samples, dispose of or recycle sample extracts, perform real-time QA/QC, and report experimental results in a format that can easily be adjusted to individual clients’ needs. The various elements of these modules will be easily interchangeable so that laboratories can maintain maximum flexibility in responding to the marketplace. These systems will be analogous to the “flexible manufacturing” systems being adopted in many industries, where tooling and other adjustments to assembly lines are accomplished through the use of computers rather than manual adjustments.

The role that computerization and automation will play in the industry is likely to vary somewhat with the size of the lab. Larger labs appear to be having a greater problem in maintaining a viable cost structure than are smaller labs and will stand to benefit to a greater extent if computers can be effectively used to lower overhead and data management costs. As discussed earlier, to obtain the full benefit of these productivity-enhancing tools, it may be necessary for the large labs in the industry to get even larger. Because industry revenue growth will probably continue to be relatively modest, the primary route for labs to become larger quickly in coming years is likely to be through mergers and acquisitions.

It is somewhat ironic that computerization and automation will be playing such a pivotal role in the restructuring of the laboratory industry, because the advent of computer-controlled instrumentation is partly responsible for the glut of analytical testing capacity that currently plagues the industry. The availability of sophisticated, relatively easy-to-use instruments (some of which are refurbished, older units to which electronic controls have been added) dramatically lowered the barriers to entry during the 1980s for small, relatively poorly capitalized competitors. The result was a flood of new competitors at about the same time that revenue growth was beginning to slow.

Opportunities will also exist in coming years for companies that provide field testing products and services. The need to reduce costs and provide maximum flexibility for clients is also driving this trend. One of the biggest complaints that customers consistently have regarding the laboratory industry is that it takes too long for labs to process analytical results. Time is literally money for these customers, because often the crews and equipment at a cleanup site have to wait for the results of the testing before proceeding with remediation.

It is unclear at this point, however, whether environmental labs will benefit or be hurt by this trend. More samples being tested in the field will translate to fewer being tested in the lab. Labs that will be most successful in adjusting to this shift will be those that identify the field testing needs of their clients and then move to provide those services. This will require, for many labs, a change in mission from being an environmental laboratory to being a provider of environmental analytical services.

The environmental testing laboratory industry is currently suffering through a period of intense price competition, declining profitability, and consolidation. This condition is likely to continue over the next several years until the imbalance between supply and demand for laboratory testing services is alleviated, either because of accelerated demand for testing services, or through additional consolidation. Even then, however, because of the relative maturity of the industry, a return to the rapid revenue growth rates and exceptional profitability of the 1980s is unlikely. Instead, the pressures of the marketplace will force environmental labs as never before to become faster, more efficient, more customer-service oriented, and more cost effective. Advances in information systems, automation, and field testing, in particular, will be among the factors that will enable labs to achieve these goals.

The lab industry certainly should not expect that new waves of environmental legislation, similar to those of the late 1970s and early 1980s, will relieve the pressure. Other than the CAA amendments, which are accelerating the development of remote chromatographic and continuous monitoring instruments, there do not appear to be any major environmental initiatives on the horizon that would stimulate demand growth and increase the pace of technological development. As was discussed above, most of the major environmental programs, including Superfund, SDWA, CWA, and RCRA, are likely to be reauthorized over the next several years. With issues such as health care and crime at the top of the national agenda, however, it is extremely unlikely that the reauthorization of any of these programs will result in sweeping new monitoring requirements. Instead, the probable focus will be on restructuring the programs so that their goals can be met at a lower total cost to society. In some cases, this may actually mean scaling back some testing requirements, particularly those affecting small communities or businesses.

The emphasis of environmental policy in the United States has shifted in recent years from controlling end-of-pipe emissions to pollution prevention, waste minimization, and control of non-point sources of pollution. This is a natural policy evolution. The most flagrant problems that the major environmental laws were originally designed to correct—such as the release of raw sewage directly into waterways and the inappropriate disposal of hazardous wastes—are now largely under control. Although this policy shift is a healthy sign for society, it is not particularly a positive development for environmental testing labs. Factories have been redesigning processes to reduce the amount of hazardous materials being used, as well as to lower the levels of pollutants being emitted. The net effect of this process, which is still ongoing, has been to reduce the amount of environmental analytical services these facilities require.

With the market for environmental analytical services likely to grow slowly and remain intensely competitive, it is instructive to examine the factors that appear critical for success in the industry and to explore how the trends identified in this report will change the competitive outlook for laboratories in the future.

Interestingly, despite the litany of hardships faced by the environmental lab industry as a whole, a number of laboratories continue to grow and thrive. What are the characteristics of these successful labs? Based on extensive management consulting experience within the industry, Wade Miller Associates believes the “model” of a successful environmental lab in today’s challenging marketplace has the following characteristics:

• a low-cost provider of analytical services

• strongly “niche” oriented

• committed to providing analytical services that are flexible, responsive, and customized for clients’ needs

• has sufficient financial resources to survive the current industry shakeout;

• committed to affect a paradigm shift toward viewing the laboratory as a production facility (rather than a research facility) with the primary product being information

• willing to examine every process that goes into producing the product (Information), with the aim of streamlining or eliminating processes that do not add significant value

One key theme that underlies all of the characteristics outlined above is that managerial issues, rather than technical issues, have become paramount in the industry. Sophisticated customers expect high-quality testing data from all labs, so technical prowess can no longer be used as a unique selling point. In addition, the quality of today’s instruments is so high, much of the artistry has been taken out of chemical analysis. Instead, customers are more focused on issues such as sample turnaround time, price, and responsiveness to special requests. All of these tasks will be made easier for labs as information management systems, automation, and field testing are further developed.

Financial strength is a key issue for environmental laboratories. The large inflows of investment capital from venture capitalists and other financial institutions into the industry during the middle and late 1980s have largely evaporated. Furthermore, the likelihood of continued restrained growth and profitability pressures will make it difficult to attract new investors to recapitalize struggling labs. The lack of outside capital further underscores the necessity for labs to improve operating efficiencies and lower costs.

There is a growing realization by government and industry alike that something must be done to reduce the spiraling costs of America’s ambitious environmental protection program. The “low-

hanging fruit” has already been picked off, and continued gains in environmental quality without huge new expenditures of resources will require environmental programs to become vastly more efficient. As part of that agenda, the environmental lab business will be forced to provide a higher quality product at a substantially lower price. The effective application of technology to improve productivity and increase flexibility will be a key factor in reaching those objectives.

In addition, over the next 10 years or so, it is likely that environmental policy in the United States will continue to evolve away from traditional regulatory methods. Command-and-control regulations increasingly will be supplanted by market-based regulatory schemes, which will provide industry with more opportunities for flexibility and innovation in responding to environmental mandates. Media-based regulations (affecting wastewater, solid waste, drinking water, etc.), will eventually give way to multimedia regulations, which emphasize minimization of the total environmental impacts of different types of industrial activities. For the typical industrial facility, these shifts probably will mean that the various permitting and compliance activities currently required will eventually be replaced by one encompassing environmental permit.

There are already halting steps being taken in this direction. In the United States, the EPA is reorganizing its enforcement division to emphasize coordination of the various regulatory activities that affect specific industries. The EPA also recently promulgated its first “cluster” rule, which regulates both air and water emissions of pulp and paper mills. This preliminary rule is not only EPA’s first rule to take a multimedia approach, but it is also the first rule that specifically mandates pollution prevention and waste minimization activities inside a manufacturing plant. In Europe and some East Asian

countries, this same process of emphasizing a systems approach to environmental regulation is being called “ecomanagement.” The European Community, in fact, is in the process of developing an “ecoaudit” scheme to implement this idea that may eventually serve as a model for a similar system in this country.

The ultimate impact of many of these new policy initiatives is years in the future and difficult to predict with certainty. Table 2 represents one potential model for the laboratory of tomorrow. One thing is certain: The environmental laboratory industry will be undergoing continuous radical change in coming years as environmental markets continue to evolve. The model laboratory of the future, therefore, is likely to be far different from that of today.