To answer that question we should look at the construction of DPRs, as dis-cussed in Chapter. 2. As is known, the DPR has a modular design. The basic functional units of DPRs are the motherboard with an ADC, a microproces-sor, various types of memory, and other accessories; digital (logic) inputs
with a photocoupler module; analog inputs based on the current and volt-age transformers module; and the output relays module. Isn’t this modular design of a DPR similar to that of a personal computer (PC)? However, there is one significant and even essential difference between PCs and DPRs: each PC module is available on the open market and you can assemble your PC with modules produced by different manufacturers from different countries.
What does it lead to? It leads to price reduction and the ability to assem-ble a PC with the best-suited modules regarding their characteristics and price. The same applies to the software. There is some universal platform (Windows) and a huge market of applications for every taste.
But, are there any significant structural differences between the PC and the DPR? In reality, there are few essential differences: it uses the same power source, and the same main module (motherboard) connected to auxiliary modules, such as the analog inputs module (a set of CT and VT with the filter and ADC) instead of the video card, the logic inputs module instead of the TV tuner, and the output relays module instead of the sound card. What is the essential difference between the software designed for multifunctional DPRs and any other software? There is no difference! So what’s the prob-lem? Why do we have now a vast number of totally noninterchangeable DPR designs, instead of a set of universal modules in the form of PCBs? Why do we have a wide variety of DPR software versions that are inconsistent even with each other?
To answer this question, let’s trace how this remarkable business works.
For example, what happens if some specific module of a specific DPR type at a given substation fails? This is what happens: since there is no market for universal modules, the user can replace the failed module only and exclusively with the same one, produced by the same manufacturer. Thus, after you spent a small fortune to purchase the DPR from one of the manufacturers, you actu-ally fall into the bondage of economic dependence on this manufacturer for the next 10–15 years, since after you have chosen one manufacturer, it no longer matters if there are other manufacturers on the market, as you can-not use their products. And the only way to get out is to pay a small for-tune one more time for a DPR from another manufacturer (and, thus, you go from one bondage to another). And what does the manufacturer in such a situation of absolute monopoly do? Right: increase the price! The price of one spare DPR module can reach almost one-third or even one-half of the price of the entire DPR! As you have nowhere to go, you pay that price. And what happens after 8–10 years of the DPR operation? Here’s what happens:
the manufacturer had already developed several new designs during that time and it has become unprofitable for the manufacturer to maintain the facilities producing spare modules for the old relays, so production has been stopped. What does the consumer do in this situation? Right: throw the old DPR into the garbage, even if there is only one faulty module (PCBs of the modern DPR are developed by the irreparable technologies), and shell out money for purchasing a new one. Who wins? Right: the same manufacturer!
But why is the manufacturer allowed to cash in on the consumer in all condi-tions? What should have been done to turn the tide in favor of the consumer?
That’s right: the DPR should be realized as a set of modules—circuit boards with universal standard dimensions and connectors, just like those for a PC, and with an integrated basic program shell compatible with the software for the given type of protection or protection kit available on the market.
Today each type of DPR has its own body which significantly differs from that of any other DPR type, and sometimes differs even from those from the same manufacturer. Today as a rule, a single DPR is installed in the relay cabinets: 3–5 units in each cabinet (see Figure 3.10). If a new DPR will be realized as a set of universal modules on the PCBs, then such sets (at least, in most cases) will not require the individual cases. Each DPR may be installed as a single horizontal section in the cabinet with the PCB guiderails, indi-vidual door, and rear wall with connectors and terminals for connecting external cables just as in PCs.
The relay cabinet should be made by special technology intended to protect its contents from electromagnetic influences. There are modern technologies (special cases, electrically conductive pads and greasing, filters, etc.), which could significantly reduce the effect of external electromagnetic emission of a wide frequency range on highly sensitive DPR equipment (see Chapter 8).
FIGURE 3.10
The modern method of mounting the DPR in relay cabinets.
The proposed tendency of development could open the DPR market to new
“players,” some of whom could produce analog input modules equipped with current and voltage transformers, and others could produce mother-boards, or software. The consumer could “build” the DPR out of the separate modules from different manufacturers, just as it is today for PCs, based on the cost and quality of these modules; and the consumer could use the same software for all its DPRs. It would answer many of the questions set out above and significantly reduce the cost of relay protection. This could also enable installing two sets of identical DPRs instead of one in order to improve reli-ability, and using the second set as backup, starting automatically upon the
“watchdog” signal of the damaged core DPR. In addition, it would eliminate the necessity of the individual power source for each DPR, and allow using one double high-capacity power source set of improved reliability for the entire cabinet. And finally, this would allow installing many service mod-ules, capable of improving DPR reliability, in the same cabinet.
Thus the relay protection maintenance would be simpler as the service staff would not need to read thick folios (Figure 3.11) about different DPRs installed in the facility and study characteristics of the software of each DPR type. In addition to easier maintenance and reducing the new DPR lead time, it would significantly reduce the percentage of errors caused by the so-called human factor. Such DPR design would also solve the problems of testing complex DPD functions (see Chapter 7).
How can the program of DPR reconstruction described above be realized in practice? The easiest way is to start in one big country with a large market and several DPR manufacturers able to closely work with each other. In our opinion, the first step in this direction should be the setting of national stan-dards containing requirements of the new type of DPRs, their software, their
FIGURE 3.11
Thick folios containing user information about DPRs.
test procedures, and so on. These standards must be developed, by a wide range of professionals, including scientists, potential DPR producers, their potential customers, and representatives of project organizations. It seems that such relay protection development, even in one country, would show an example to be emulated around the world.
3.5 “Intellectualization” of Protective Relays:
Good Intentions or the Road to Hell?
There was an interesting story in an old science fiction novel. It all started from such an innocent thing as an odd night call to all the phones on planet Earth. The call announced the birth of Global Intelligence to all people of the Earth. It turned out that at some stage of development, the proliferation of computers escalated into a new quality: the millions of computers, inte-grated into a single network that controlled everything and everyone on planet Earth, suddenly became conscious of themselves as an entity capable of reproducing itself through the automated factories and robots connected to the same network, as well as capable of defending itself with computerized weapon systems designed to destroy the human race. From the perspective of Global Intelligence, humanity was nothing but a useless vestige, gobbling up the planet’s resources. There are no prizes for guessing about the further development of the action.
Network-connected computers already control almost all types of modern industrial production systems, water supply and electricity systems, telecom-munications systems, and networks. New terms such as smart grid and relay protection with artificial intelligence have emerged in technical literature rather than in science fiction. Today technical literature rather than science fiction refers to the creation of a so-called smart house, where even the refrigerator will analyze the stored products, and, based on the analysis of consump-tion, will make an order and send it over the network to the nearest super-market. Today you can find microprocessors everywhere, even on the water closet lid.46 Humanity is moving by leaps and bounds to the creation of powerful Global Intelligence prophesized in the old science-fiction novel.
But let’s get back to reality. And the reality is that major failures in the energy systems that have occurred in America and Europe (the United States in 1965, 1977, and 2003; France in 1978; Canada in 1982 and 2003; Italy in 2003;
the United Kingdom in 2003; and Sweden in 1983 and 2003) were caused by incorrect or, rather, unpredictable actions of relay protection during complex emergency modes due to disabling the wrong sectors of the network in every particular situation. Had the action of relay protection under these specific circumstances been different, the system failures might have been avoided.
I am referring to the power systems (in the United States and Western Europe)
that have already been equipped with computers and microprocessor-based protection. For comparison, I should mention that one of the world’s largest power systems with a negligible percentage of computerized relay protection devices and the old worn-out equipment has never suffered from such fail-ures. I am referring to the power system of Russia. The answer to the ques-tion about the reasons for this can be found in the book of E. M. Schneerson,29 foremost authority in the field of modern relay protection:
Improvement of the technical level of relay protection devices (RPD) alone does not necessarily lead to the equivalent improvement of effi-ciency as related to the response to emerging damages. For example, outdated electromechanical and to an extent electronic static RPD, if pro-tection functions and settings are chosen correctly, certainly provides better protection for the network than the microprocessor RPD without rational definitions of the specified parameters.
In fact, the behavior of electromechanical and electrostatic RPDs under emergency situations was rigidly determined by their operating principle and settings. Current trends47–52 in the development of DPRs are associated with the increase of their “independence” (that is, in fact, unpredictability) in making decisions. It relates to the relay protection self-learning capability peculiar to adaptive neural networks, as well as the use of technologies of artificial intelligence with fuzzy logic, and the like.
Another clear trend in the development of the modern DPR is the excessive complexity by including extraneous functions not typical for relay protec-tion. Here, for example, is the list of functions performed by the so-called intelligent controller (Intelligent Protection and Automation Controller iPAC by Dynatrol Systems Inc.):
IEEE protective functions (ANSI number):
Sync Check (25)
Inverse Time Over Current (51)
•
Under- and Over Frequency (81)
•
Lockout (86)
•
Differential for Transformer Protection (87)
•
Measurement functions:
Voltage and Current RMS Values
•
Neutral Current RMS Values
•
Power Factor Measurement (power factor correction: capacitor
•
bank switching)
Total Power Measurement
•
Real and Reactive Power Measurement
•
Power Quality Measurement (FFT for harmonic measurement)
•
Coil Monitoring for relay failure detection
•
Cold Load Pickup Logic to prevent protective devices from
oper-•
ating when cold load is put on the circuit
Voltage Constraint with Current Pickup lowered to increase
sen-•
sitivity when voltage is also collapsing during the fault
Breaker Control Blocking for coordination with upstream
•
and downstream protective devices via DI or peer-to-peer communication
Directional on overcurrent devices
•
Add to this the monitoring of external current and voltage circuits, the reg-istration of events, the functions of an emergency digital oscilloscope, and other routine functions of the DPR. Then there is the danger of excessive concentration of protective functions in a single terminal (e.g., a micropro-cessor relay-type SYMAP by Stucke Elektronik GmbH incorporates 39 sepa-rate protection functions); additional relay protection functions, extrinsic to the protection itself, lead not only to the physical complication of the device, consequently reducing its reliability, but also to the complication of its soft-ware and user interface. This in turn leads to a sharp increase in the number of software errors (again, the so-called human factor). Due to such a large number of functions, using the same internal resources of DPRs and pos-sible conflicts of the embedded logic functions during complex emergency mode accompanied by a transition of one type of damage to another, it is not always possible to predict the behavior of the protection. Damage to one function that is common to all the internal-element DPR functions (power supply, watchdog, memory, microprocessor, its servicing subassemblies, etc.) will result in the instant failure of all protective functions at once.
Despite the obvious problems existing today due to the excessive concen-tration of protective functions in a single terminal, some leading experts in their philosophizing about the future of relay protection not only advocate additional “adding on” of extraneous functions to relay protection, but also
even go further. They put forward the fantastic idea of “multi-dimensional relay protection”50 and “relay protection with proactive functions,”52 acting on the basis of its own experience, its own analysis of the current status of the protected object and prediction of its future state. In essence, this is about the relay protection capability taking completely unpredictable actions, as an independent intelligence making its own (previously not determined) solu-tions and changing power systems operating modes (through switches) at its own discretion before the emergency mode occurs.53
It must be emphasized that there is nothing wrong with the development of computer-based diagnostic and prediction methods for electrical equip-ment, and it could only be welcomed but for the attempts to “intercross” it with the protection relay.
Apart from the risk of losing control over the relay protection actions, cur-rent trends of its development dramatically increase the risk of hacker attacks on the grid because a computerized relay protection system is a good target for changing the state and affecting the modes of power systems. Despite the serious concern of specialists about this problem,54,55 the trend toward greater susceptibility of power systems to hacker attacks only grows.
Another serious threat to the stability of power systems, which is based on current trends in its development, is the development of technologies of artificially produced destructive impacts on electronic and computer equip-ment.56–61 The development of these technologies throughout the world contributes to increasing the spread of microprocessor technology and the memory elements with high sensitivity to external electromagnetic emis-sion, on the one hand, and the tendency of constantly increasing the density of microelectronic components as a result of the reduction of the thickness of the operating and insulating layers in the crystals, on the other hand. These two tendencies, directed toward each other, form a very dangerous vector of development of modern technologies.
Moreover, today you do not need any special knowledge or equipment in order to create a device capable of destroying all the electronic devices of your neighbor. You can find numerous descriptions on the Internet of such devices based on common microwave ovens (see Figure 3.12).61
As for the special “combat” units, there have been very impressive achieve-ments in this field (see Chapter 8). Compact mobile ultra-broadband pulse generators with a capacity of 1 billion watts, portable emitters in the form of pistols and rifles, explosive devices in the form of attaché cases destroying all the electronics within a radius of many hundred meters, special ammuni-tion, and other weapons are designed specifically for the remote destruction of electronic equipment.
So, where are we going, and where will we arrive? Why do the current trends go unnoticed by specialists? Obviously, there are a lot of parties very much wanting this trend to continue. However, in our opinion, it is not a question of whom to blame, but of what to do.
In contrast to isolated measuring and monitoring of computer systems, the protective relay is associated directly with the possibility of destructive impact on power system modes. This is the most important and fundamental difference of relay protection from all the other computerized devices and systems used in electric power engineering, preconditioning a need for a different approach to relay protection.
The above thesis, in our opinion, is the answer to this question. What’s needed is a different approach in maximizing the reliability of relay protec-tion and avoiding features unrelated to relay protecprotec-tion, limiting the number of functions in a single microprocessor terminal, avoiding algorithms with nondeterministic logic allowing the unpredictable action of relay protec-tion, taking advantage of maximum simplification of the user interface, and conducting special research and development providing for the operation of relay protection against intentional destructive electromagnetic influences, for example by introducing stand-by emergency relay protection sets. Only electromechanical relays resistant to intentional electromagnetic influences, requiring no live power, and thus being always ready to work can be used as such stand-by relay protection sets. Therefore, in our opinion, it is too early to dismiss electromechanical relays. Rather, they should be improved through new technologies and materials, and their range should be updated.
Since the relay protection algorithms are not so complicated (all of them were effectively realized with electromechanical devices, which now account for over 90% of all protective relays in Russia), the current protection devices can be as simple as possible. There have been no new functions introduced in relay protection by DPRs, but only some relay protection characteristics were improved. In particular, distance protection received polygonal characteris-tics instead of the circular ones of old electromechanical relays. Therefore, in
FIGURE 3.12
Electromagnetic weapons based on a household microwave oven.
reality there are no objective reasons for today’s substantial complication of the relay protection functions.
On the other hand, recently more and more complicated and sophisticated systems for monitoring electrical equipment modes based on the continuous
On the other hand, recently more and more complicated and sophisticated systems for monitoring electrical equipment modes based on the continuous