The first evolutionary step from these localized metering systems was to provide remote readability. Industry needed metering capability in order to check the utility and monitor their own capacity for switching and starting high power electric motors. The evolutionary development of power monitoring and control systems for utility and industry has continued to unfold independently. Looking back in time from today's deregulated perspective, the difference between utility and industrial power monitoring and control systems are artifacts of the boundary created by the installation of a revenue meter.
Not all of the functional differences in systems evolution are artifacts of regulation. A number of real functional requirements have driven the development of power monitoring and control systems for industrial and utility use. In one sense the basic instrumentation for both is measuring the same physical parameters, but the monitoring of the parameters is being done for different purposes.
In the industrial situation, most of the functional requirements are aimed at economics, i.e., reliability and uptime and allocating electricity usage costs to areas of the industrial facility and/or product lines so that the electrical system function is not an impediment to production.
On the utility side, uptime again is the objective, but the approach is more proactive than reactive. Dynamically managing loads and distribution systems against generation and transmission systems is of a critical nature. For utilities, remote reporting over a large geographic area is a necessity, however, industrial sites are often large enough to have similar requirements. In fact, the industrial systems in many cases have remote reporting that spans the entire service area of a utility, particularly when energy usage and costs are being aggregated for accounting purposes in large multinational organizations. Thus the factors driving the development of industrial power monitoring and control systems, or monitoring exclusive of control, were and still are, reduction in labor and timely recording of meter readings.
In industry, minicomputers made by companies such as Digital and Data General among others, were being employed by the late 1970s in relatively elaborate power monitoring and control systems. Using analog metering, they were deployed in industrial processes that were both energy intensive and business critical. Much of the drive for the development and installation of these systems was probably rooted in the reaction and paranoia of the engineering managers who had survived the recent oil crisis. By the early 1980s some extremely large and complex custom proprietary systems for power monitoring and control had been developed by the chemical and refining industry. Many of those systems had development costs for software alone in the $1 to $3 million range, as measured in 1980 dollars. Overall systems cost, including hardware, software and installation, was in the range of $0.5 to $1.5 million. At his price, the applications for which such systems were viable were extremely limited. Management enthusiasm for broader deployment of this technology waned, as the memory of the oil crisis faded and the price of a barrel of oil fell through the floor.
At the same time, minicomputers were being widely applied to utility concerns in the creation of centralized generation, transmission and distribution management. Because of the nature of the technology, these system architectures were devices of centralized control, with long and far-reaching tentacles for scanning relatively large numbers of metering points over a wide geographic area. >>From a vendor perspective, systems development was carried on by mutually exclusive teams, and for the most part by mutually exclusive companies. Where it was done within the same company, there was very little discussion or dialog among the teams.
GENERATIONS OF PM&C SYSTEMS TECHNOLOGY
For the purposes of this paper, we will begin our review of technology from the industrial perspective, since in most cases this covers technology which is considered to be commercial.
First Generation PC Based Systems
Earlier this year in a commercial technology review of power monitoring and supervisory control systems technology, I outlined the development of two generations of systems technology[1]. The first generation was PC / DOS based. A second generation, now coming to the market, is mostly based on the NT system.
By the late 1980s, several small vendors were beginning to offer remote power monitoring systems with PCs running extensions to the DOS operating system. These early PC-based systems, far lower in cost than their predecessors, had multi-tasking capability. They were sufficient for scanning a serial line running a multi-drop protocol to field instruments, which in turn were running either proprietary of PLC subsets and maintaining some level of operator display, data logging and query support. Their creation and the market for them, was driven by large industrial facilities that were not necessarily power intensive but whose physical scale and size made remote meter reading a viable alternative to union wage labor carrying clipboards. Perhaps, more important was the need for systems with some SCADA-like functionality for the recently empowered cogenerators. They were discovering the potential value in selling power back through the utilities, now required by the Public Utilities Regulatory Policies Act of 1978 (PURPA) to purchase for them.
Second Generation Systems
Early 1996 saw the commercial beginnings of a new generation of industrial power monitoring and control systems, driven to market by three factors:
The second generation systems have begun to divide into two tiers, a monitoring system focused on remote metering, and a second tier which is a modular system capable of fully implementing both power monitoring and supervisory control. It also includes some capability for direct distributed control within the individual metering units. The vendor jury still seems to be out as to whether power monitoring and control will continue to be a different type of automation than any other industrial automation and thus sustain a need for significant specialization.
Utility systems have followed a similar evolution, but theirs has been done in the context of larger information systems, frequently referred to as energy management systems. While operation engineers have been concerned with substation distribution requirements, utility resource planners have been concerned with transmission systems and interconnects. While revenue managers have been concerned with the expense of residential meter reading, utility managers have been concerned with everything from the expense of residential meter reading to economic evaluations and planning centered around cost of kilowatts sold vs. cost of actual generation of these kilowatts.
Toward the Future
As we approach the year 2000, the functional requirements for power monitoring and control continue to expand, the functional requirements between industry and utilities do not converge. On the industrial side the evolution of power monitoring and control systems has been driven by technology cost/performance -- there was a demand for a solution, once the price became right. We think this driving force has probably run out of energy. For utilities, power monitoring and control systems are the future. In a word in which electric power is becoming a commodity, bought and sold dynamically, the diversity will cease expanding. The ability to monitor and control the delivery process will begin the independent evolution of industrial vs. utility power monitoring and control systems to a close.
By the end of this century, requirements for systems may become more unique, distinct and clear, but their functional requirements, when looked at as a continuum, will dovetail and fit together as the operational controls for implementing a dynamic electric marketplace. We see the driving factors as:
COMMERCIAL INDUSTRIAL TECHNOLOGY
In preparation for our survey of power monitoring and control systems we reviewed a number of commercial systems vendors, including ABB, Allen-Bradley, Bailey, Cutler Hammer/-Westinghouse, E-Mon, GE, Honeywell, Jomitek, Multilin, Operation Techno-logy, Power Measurement, Scientific Columbus, Siemens and Square D. We then developed a generic reference model that could be used to describe most of the commercial technology available today.
Our reference model diagrammed the major PM&C system functions of metering, monitoring and control, with the information flows and subcomponent functionality shown at the next level of detail. We logically separate metering from monitoring and control because metering is frequently a functional component that has no interdependencies with control. In some physical applications, however, control may be packaged with metering and control algorithms may repetitively resample metering, but this does not in any way affect information flow in the ongoing operation. Thus our logical model is valid for both a local control unit, with metering, monitoring and control, or a distributed control system where metering, monitoring and control may be functionally separated.
Comparison of PM&C Systems:
Industrial and Utility Applications
Before looking at the system archetypes for industrial technology in detail, it is useful to look at a block diagram of utility logical system functionality. The major functions are data acquisition, supervisory control and alarm display and control. As you can see from comparing the differences between the industrial model and the utility model, the context in which these systems are used significantly affects how very similar functionality has been integrated in very different ways to meet their individual and unique needs.
From a high-level systems view utility and industrial systems can be thought of as two systems in parallel, completing the end-to-end connection from generation to consumption. From a vertical perspective, the systems perform a large number of overlapping functions, in that they both must collect data, monitor and consolidate information, and provide controls appropriate to their sphere of influence. This conceptual view of PM&C systems for both industrial and utility applications will become the dominant view as the technical ramifications of utility deregulation unfold.
In tribute to the historical artifact of separation, we note that industrial systems have primarily been viewed as standalone systems. They may have weak links of data interfaces to process control, shop floor control or other factory information systems, including production scheduling, MRP (Material Requirements Planning) of ERP (Enter-prise Requirement Planning).
In the utility environment, monitoring and control systems are often much more tightly integrated with the utility's broader management system. Because of the physically distributed nature of these systems, they are frequently referred to by the name SCADA (Supervisory Control and Data Acquisition) and are usually interfaced with IMS systems (Information Management Systems) and UI (User Interface) systems. SCADA could as well be applied to industrial systems but usually is not.
Utility EMS systems should not be confused with the large-scale minicomputer and industrial energy management systems of the early 1980s, which were also used to monitor high energy consumption or cogeneration applications. In their architecture and structure they looked nothing at all like utility systems.
In the regulated world, the metering point not only created a boundary between the utility and the customer, but also an imaginary wall in the evolution of power monitoring and control systems technology, specifically its application before or after the meter. In fact, in many of the supplier companies, the systems technology was evolved and developed by separate engineering groups, marketed by separate marketing organizations, and in many cases did not even use common internal components.
Functionality and Purpose
As mentioned earlier, second generation industrial systems divided themselves into two functionality classes, the first being entry level monitoring systems that evolved from mechanical meters and recorders. The purpose of these systems was to monitor for reliability. The focus was on providing numerous benefits to the operating engineer, whereas in the world of the utility the operating engineers rarely concerned themselves with the billing meter.
The development of the industrial monitoring system, from the beginning, was driven by labor reduction as a motivation. In contrast, the utilities for years have used many meter readers, all walking miles, and are only now beginning to recognize monitoring systems as potential labor cost savings. In terms of supervisory control, industry leans more toward automation while utility systems seek to provide more centralized readout for operator intervention in control centers. Although hardware computer configurations were fairly similar, the functionality and ultimate purpose for which the information was used drove the focus of development.
Systems Architectures
Historically, systems architectures were somewhat different because of the limitations in telecommunications technology. While industrial technology in the past could be classified as either embedded or not embedded, master-slave, or hierarchical, the look toward the future appears to be NT, the label for the Microsoft version of an industrial operating system. It is our belief that as vendors begin to benefit from this new platform, there will be a significant shift in commercial technology, affecting both the industrial and the utility marketplace.
For utilities, modern and telephone technology, radio modem, and other telecommunications devices that capitalize on the recently opened radio spectra will begin to provide opportunities that were unthinkable in the regulated world.
An additional factor contributing to this shift and shakeout is the sophistication of independent intelligent devices. Many of these are now fully network compatible with plant and information systems networks, unlike the slower and more primitive point-to-point and multipoint technologies of RS485. While we see the advantages of the more computer-based systems, we recognize that their use places an increased responsibility on the application engineer. Power monitoring system independence may be sacrificed, with potential loss of the robustness, reliability and fault tolerance that could be maintained in an entirely independent power monitoring and control system with an isolated independent communications network.
As utilities become more consumer-focused, understanding the industrial application of PM&C technology will become more critical to their business success. As one can see, the interface between the functional requirements is clearly end to end.
SURVEY OF PM&C SYSTEMS AND THEIR APPLICATION IN USE
In late 1996 ITR conducted an extensive survey of PM&C systems and their application in use, covering company demographics, energy supplier perceptions, power monitoring and control systems technology, and computer platform and communications for the systems. The survey also included questions on overall usability and satisfaction with systems technology[2]. Responses were received to the 100+ question survey from 70 sites: 73 per cent in businesses with over $1 billion annual revenue, 13 per cent between $500 million and $1 billion, and the balance, 14 per cent, between $50 million and $500 million. Approximately 65 per cent of the sites surveyed were manufacturing, 19 per cent classified their businesses as mining operations, 9 per cent transportation and utilities, with the balance spread across services, printing, construction and commercial facilities.
We found that energy is chiefly managed at the site level. Further description of the survey participants shows that 57 per cent were energy consumers only and the balance had some cogeneration capability on site or within their organization, accessible to their site. A number of respondents had more than one system on their sites and in many cases had more than one vendor's technology. Responses to the survey showed a representation of 14 different vendors.
System Functionality and Usage
Survey respondents were asked to identify the system functionality most used by their site. Distribution of the responses is shown in Figure 3. In general, the survey results indicate that there is clearly a break between the core functionalities of metering, power factor, correction equipment, and utility monitoring, as distinct from control functions used for energy management such as load shedding and peak shaving. It is worthy of note that a significant number of respondents indicated that they used their systems for spot market buying and purchasing under variable price structures. Future survey work should perhaps focus on these areas.
We tried to gain some indication from the user's perspective about user satisfaction. Most users, 47 per cent, indicated that they would purchase the same system again. Changes in vendor selection seem to be driven more by the need for functionality not provided by the current vendor, or when another vendor has leapfrogged in technology and is producing superior technology. Maintenance, reliability and cost were less significant.
Platforms and Communications
Respondents were asked to indicate the computer platform used for their currently installed PM&C system, as well as their preference for future platforms. In the installed base, the largest number were reported to be propriety hardware and software. We conclude that the MS-DOS platform in past years attracted a large number of users and still amounts to 25 per cent of the installed base. The PC MS-DOS systems may have been larger when we notice that the Windows 95 and NT systems collectively total nearly another 25 per cent. Of significance is the fact that Windows NT in this survey has 50 per cent more usage than Windows 95, even as of a 1996 survey. When the same question was asked about platform preference for future systems, the most significant drop was in proprietary, with NT and Windows 95 nearly doubling, UNIX making a slight gain and VAX-VMS holding its own.
In terms of installed networking technology, fiber optics\ has become the predominant choice with 47 per cent of the total, and base band, mostly a traditional Ethernet, second with 14 per cent. Broad band and radio modem technology also showed measurable presence, each having an equal share, 8 per cent, in the distribution of the pie.
Satisfaction and Usability
Users were asked to rate their currently installed PM&C system relative to a number of criteria. We asked them to rate these criteria on a scale of 1 to 5, where 1 = outstanding and 5 = unsatisfactory. Figure 5 requires considerable study to absorb, but to highlight the extremes, ability to obtain necessary data received the lowest number of the respondents who rated their system as outstanding. Overall, the highest number of respondents gave their system an unsatisfactory rating in the area of end user customization ability.
In response to a question focused on understanding who uses what information produced by the PM&C system, we found that system usage by function was reasonably well grouped by professional job description. We found that at the procurement and cost control end of the spectrum there was high value from accounting and maintenance, with production and engineering focused more on uptime and reliability. While there was commonality in need for systems status, other functions, such as reliability in production scheduling, were only important to one type of job and operation.
VENDOR SELECTION
Users were asked two survey questions related to the criteria they used for selecting their currently installed PM&C system. A summary of their responses, where higher values indicate the more important selection criteria. Again, responses of a higher value indicate more importance. In comparing the two figures there is little difference from a technology viewpoint, but a significant increase in value for allied vendor functions, such as training and documentation, in support of systems quality.
System Justification
One part of the survey focused on understanding how systems projects were justified. When asked if a financial justification was developed prior to the installation of the system, the majority of respondents, 53 per cent, said yes. Of the remainder, 29 per cent were sure that no justification had been prepared and 18 per cent were uncertain whether any justification existed at all. When asked more specifically if any evaluation of the value received, i.e., financial results, had been developed since the system was installed, the majority, 57 per cent, said no. The second largest per centage, 27 per cent, was unsure, and only 16 per cent indicated that any post-installation financial check on the project had ever been done. Responses to a later question, asking if the systems had met financial objectives for which they were initially justified, led to some interesting inconsistencies. A near majority, 45 per cent, thought that the systems either met or exceeded the objectives, even though earlier answers indicated that no financial analysis had been done.
Upon identification of this inconsistency, further work was done to compare these answers with the demographics and qualifications of the survey respondents. In the majority of cases, the respondents providing the seemingly inconsistent answers were represented to be qualified engineers, mostly senior level, and almost always with some number of years of experience in the area of the works being surveyed.
This section of the survey points out the need for engineers to work toward a better understanding of the implications of their projects, and the results and benefits from their implementation. While somewhat surprised at the survey data related to system justification, it is consistent with our experience in our consulting practice. A frequent concern of industrial facilities managers is their inability to justify power monitoring and control systems technology in spite of their intuitive feeling that these systems are an operational imperative.
CONCLUSIONS
The power engineering community has some lessons to learn from their counterparts in the process industries. PM&C systems for industrial processes are designed to manage the production and delivery of commodities. As we move toward a deregulated world, electricity clearly fits that category. When we begin to look at the difference in industrial and utility power monitoring and control systems vs. those of the chemical industry for example, we find little SPC (Statistical Process Control) for quality purposes. Statistical tools are noticeably absent. Predictive and adaptive control outside the generation facility is nonexistent. Short interval scheduling, production optimization, optimized routing and individual customer satisfaction information for the utilities is not present, nor is alternative buying source information for optional delivery mechanisms for industry. Typically, all these features are present in the PM&C systems technology and allied systems that are used in other commodity industries.
The commercial technology that is available to today's industrial power engineers is usually more than sufficient to the task at hand. The issue they face is not one of technology but of linking the technology to business management. Perhaps the utilities have something to contribute in this area. As both utilities and industry begin to use this technology, the prospect is that the invisible metering point will truly stay that way. Integrated systems that cross this imaginary boundary will be able to incorporate function and facility around power quality, power delivery and energy usage optimization functions. These factors will be perceived as value adding by the customer, and as differentiating marketing characteristics by the power supplier.
As industry begins to understand that electricity, once viewed as a non-competitive element when it was priced equally to all industrial producers under regulation, will become a competitive variable in the cost equation, it must be monitored and controlled like any other significant process variable. Utilities will have a role to play in steering the industrial systems vendors as they try to bridge this gap, tie together the two worlds, and begin to educate the end-user based on how an integrated process monitoring and control system, so well known to the process industry manager, could be beneficial to the electric power site manager and the electric power marketing and delivery organizations.
References
[1] Commercial Technology Review - Power Monitoring and Supervisory Control Systems Technology, J. Becnel, K. Nicholson, M. Washburn; Integrated Technology Research, a Division of Syzegph Corporation, 1997.
[2] A Survey of the Application in Use of Power Monitoring and Supervisory Control Systems Technology, K. Nicholson, J. Becnel; Integrated Technology Research, a Division of Syzegph Corporation, 1996.
[3] "The Electric Light in Houses", Harper's Weekly, Volume XXVI, No. 1331 (1882).
Kenneth Nicholson is with Syzegph Corporation. Article reprinted with permission, 1998 Syzegph Corporation, PO Box 1790, Wilmington, DE, 19899-1790.