canadian puregas equipment limited
Proactive Maintenance and Attributes of a Technologically Advanced Outside Plant Monitoring System By Ingolf Plath Copyright June 02,1996 Canadian Puregas Equipment Limited Table of Contents 1. Introduction 2. Pressurization 2.1. Pressure and Flow 2.2. Leak Locating 3. Maintenance 3.1. Types of Maintenance 3.2. Monitoring and Proactive Maintenance 4. Outside Plant Automation 4.1. Transducers 4.2. Data Acquisition Equipment 5. Software 5.1. Operator Interface 5.2. Outside Plant Database 5.3. Flexibility 6. The Complete System 6.1. Alarms and Communication 6.2. Leak Locating 6.3. Allocating for Proactive Maintenance 7. Summary 1. Introduction Pressurization of cable was first instituted in the 1940's to protect telephone cables from moisture intrusion and subsequent problems due to sheath and/or splice integrity faults. A dry pure gas such as air was forced, under pressure, into the cables and was simply monitored by the flow rate at the point of insertion or the pressure in the cable. A technician would manually take a reading at the flowmeter or pressure test valve to discern if system integrity had been compromised. Over the years, with advanced technology and automatic monitoring, leak detecting and locating systems were deployed. The newest systems are computerized and fully automatic with customizable features. This discourse shall provide an understanding of the basics of a good monitoring system and why a monitoring system utilizing the newest technologies and techniques is important for proactive maintenance. This is not a complete in- depth analysis on pressurization, but rather a primer on monitoring systems to help those persons who require information to understand the benefits of the newest technology and how it can improve the bottom line. 2. Pressurization 2.1. Pressure and Flow Pressurization of cables is done to preclude moisture from entering within the sheath of the cable by way of natural humidity and water by way of sheath faults and susceptible splices and terminations. The ideal outside plant (cables) will have no leaks, a constant air pressure and low air flow. Due to the actual integrity of systems and their inherent leaks many systems were built with air pipe distribution so that the air can be distributed to extended sections of the cable. Otherwise these sections would have no flow or pressure as the air introduced at the beginning of the cable would have exited the cable through the leaks. One of the reasons for this low cable pressure integrity is the lack of complete monitoring and proactive maintenance. With a good monitoring system and proactive maintenance pressurization of the outside plant can be efficiently achieved with a good quality air source, properly sized and minimal air feeder pipe or none at all. When cables were originally pressurized from nitrogen bottles, and if there were no leaks at all, the pressure in the cable would stabilize. Once the pressure is stabilized then there is zero flow. Since it is almost impossible to have a completely airtight system (and there are reasons not to) we always have air flow into our cables. Air is introduced at the beginning of a cable at the central office (C.O.) which has a certain flow and pressure. As long as the flow is constant the pressure in the cable will stay constant. A cable has a certain pneumatic resistance, due to wire pairs or fibers within the cable, which affects the flow of air along the cable. The flow in a cable decreases in relation to cable length and leaks. Pressure decreases in relation to leaks and constant flow. If a high flow is registered at the beginning of a cable this indicates a leak. In the first section of the cable the pressure will not change dramatically as long as the air source can increase the air flow to compensate for the leak. The further distant from the air source the leak is, the longer it takes for the flow to increase due to the pneumatic resistance of the cable. At the further distances the pressure will drop and indicate a leak problem because the flow of air from the air source will increase slowly. Monitoring flow at the air source is important as it will indicate problems at the first section of cable. Monitoring pressure along the cable will give early indications of problems along the whole length. Both should always be monitored for the best indication of outside plant integrity. Leak locating using flow readings only gives a position between two reading points along the cable. The use of flow and pressure is a more accurate way of leak locating. The most accurate leak location method is based on accurate pressure readings and calculated by a computer using a mathematical algorithm based on a parabolic regression formula. 2.2. Leak Locating Leak locating by manual methods has progressed from visual (bubble test) through helium introduction and ultrasonic locating to leak calculating and graphing from field data of pressure and flow. The most common type of leak locating calculations are cumbersome and require expertise. Manual calculations also inherit inaccuracies from data that is generalized, such as cable resistance, flow and pressure readings and non linear instruments and transducers. The final outcome of any leak calculation cannot be more accurate than all the tolerances of any inputs combined. In other words, any tolerance introduced at the beginning of a calculation expands the tolerance of the result accordingly. When all the beginning tolerances are factored into the calculation the final result is therefore more ambiguous than we would like and this in turn costs time and money when leak locating. This then leads to reactive maintenance (emergency calls) due to time and cost constraints. A computerized monitoring system with an accurate leak location method eliminates this tedious procedure. When information from the field transducers is received at the occurrence of an alarm or interrogation, the leak location can be immediately discerned and repairs completed. 3. Maintenance 3.1. Types of Maintenance There are three types of maintenance that can be performed to maintain a plant and correct breakdowns. They are reactive, preventative and predictive maintenance. Reactive or emergency maintenance is a repair that is done after a breakdown has occurred. This is usually caused by an external or unforeseen cause. The causes can include sheath penetration, storms or equipment failure. Too often the main cause is due to a lack of proper routine maintenance because of a shortage of time, manpower or skills. This is the most costly and inefficient way to maintain a plant. It is analogous to not fixing the hole in the roof of your house until it is raining, in which case the repair becomes more costly and leads to more damage and complaints from the customer, i.e. your spouse. Companies that rely only on reactive maintenance procedures in a highly competitive market usually lose their customers and their maintenance costs soar until they become insolvent. Preventative or preemptive maintenance is scheduled or routine maintenance. This is the most common form of maintenance which is based on the equipment manufacturer's or industry recommendations. It is considered proactive. This scheduling of maintenance is fundamentally to prevent emergency repairs by improving or repairing equipment before failure occurs. Preventative maintenance has a structured and known cost, but due to the nature of a cable plant and myriad external influences emergencies will not be eliminated. There will always be unscheduled breakdowns. Predictive maintenance is unscheduled maintenance based on data collected from the actual plant. It also falls under the proactive maintenance category. From the data collected we can see the condition of the equipment and plant and base our maintenance procedure upon this condition. Monitoring of the plant provides us real time information so that we can conduct our maintenance accordingly and greatly reduce our maintenance and downtime costs. The three types of maintenance can be compared to the maintenance of a roof. If you replace the roof covering of your house only when it leaks so thoroughly that you have water damage this would be considered emergency maintenance. If you replace your roof when the manufacturer recommends this is considered preemptive maintenance, even though you may have problems before this time or the roof covering may have an actual longer life span. If you monitor or inspect your roof twice a year and repair it before it leaks, be it before or after the recommendation of the manufacturer, you have performed proactive maintenance and preempted any emergency and ancillary damage, and received the useful active life of the product. Your cost of monitoring is minimal compared to damage costs of emergencies and the cost of unwarranted replacement. 3.2. Monitoring and Proactive Maintenance In today's competitive market it is imperative to have your system operating at peak efficiency without downtime. Predictive maintenance (monitoring) is completely proactive. A proactive maintenance program produces long-term gains if properly implemented. Part of this implementation includes the purchase and use of tools to monitor the plant and its trend. Fully effective preventative maintenance requires a disciplined and well structured maintenance management program whereas predictive maintenance relies upon the monitoring system and trained personnel. Most proactive maintenance programs require a significant corporate investment in time and resources for implementation. Monitoring systems also require corporate investment and time for implementation. A good monitoring system will have a higher initial cost with earlier results from this investment. Many companies tend to be penny wise and pound foolish with maintenance programs which inhibits proactive maintenance. There are few short-term gains. It is an investment for long term gains and when fully and correctly implemented those gains are substantial. Predictive maintenance can deliver benefits to your company on many levels. Studies have proven that predictive maintenance on average is one third the cost of reactive maintenance due to continuous savings. For an example of savings calculate the cost of a repair when a cable fault is accurately located by a monitoring system and repaired before any problems arise to the cost of repair and downtime if the cable became wet and failed. These costs can be measured in actual dollar figures. As a direct result of eliminating emergencies the tangible effects of any revenue loss can be determined. By using monitoring and a predictive maintenance program, corporations willing to invest resources will gain real tangible benefits. In the communications industry the outside plant is the factory that produces revenue. The foundation (production machine) of the plant is the cable. If the cable (machine) is maintained there is no factory downtime and loss of revenue. Other benefits are: savings in personnel and time, efficient use of ancillary equipment such as air dryers, improvement of the overall efficiency of the outside plant and customer satisfaction. 4. Outside Plant Automation 4.1. Transducers A proactive maintenance program requires a monitoring system which gathers data from different points in the outside plant. The devices which take measurements at these points are transducers. A transducer changes a physical measurement into a signal readable by the monitoring system. A simple pressure gauge can be considered a transducer. The gauge physically measures pressure and mechanically changes (transduces) that pressure reading into a visual signal that we read on the dial face. Originally air pressure problems on a cable were signaled by a pressure contactor which was set at a certain pressure. The contactor would signal if the pressure dropped below the set point, but would not give a pressure reading. The technician could then use helium to detect the leak. Further developments produced the resistive step transducer that could give approximate air pressure readings. The pressure is equated to 20 steps of resistance over the full scale of the transducer. Flow transducers were also based on this same principal. Accuracy was restricted to the size of the step and the quality of the transducer. Since these readings are inaccurate helium and ultrasonic detection equipment are still widely used. Calculations based on such readings tend to be skewed. Another inherent problem is that each transducer utilizes a pair of wires which are then not available for subscribers. The latest technology pressure transducers are completely solid state and pressure readings are linear. Solid state transducers produce extremely accurate readings. They are also addressable so that up to 127 transducers can be placed on one pair of wires. The best transducers are completely sealed and tested for helium intrusion because if helium is used on a cable and displaces the air within the transducer the pressure readings will be incorrect due to the difference in specific gravity of the two gases. Pressure transducers should have a relatively small size so as to fit neatly into a splice case. Some transducers also convert the pressure reading into frequency where one herz of frequency equals one millibar of pressure. In this way it is easy to field test transducers as the frequency signal is directly related to the pressure reading. Addressing of the transducers is usually done by jumpers on contacts or by breaking contacts. Problems can arise from the former if a contact loses its integrity or in the latter a transducer cannot be re-addressed to all positions. Although these problems are not common, to overcome these drawbacks electronically addressable transducers are now available. Addressable linear flow transducers are also available with good accuracy. These can be for field use or in the C.O. Flow transducers used in the C.O. are available with visually readable scales, either bar graph or digital LED readout or both. Used along with pressure transducers, they will give excellent remote and automatic monitoring ability for proactive maintenance. On non-pressurized cable (fibre or copper) addressable humidity transducers can be used in the splice cases to monitor humidity within the cable. Transducers of this type should specifically monitor the humidity and not moisture. Some transducers available have a water sensing tape or wire. These type of sensors usually require water in the form of moisture (water droplets) for sensing. Since moisture is already within the cable for this type of sensor to activate, damage to the cable may have already commenced. Therefore humidity sensing is crucial for protection. When the humidity of the air inside a cable rises we know we have an intrusion of outside air and proactive maintenance can be commenced. We can also combine a humidity and temperature transducer in the same splice due to the addressability which enables reference to the actual dew point of the air. Other types of addressable transducers available detect temperature, humidity, current, voltage and water for monitoring the C.O. and any ancillary equipment. Addressable resistance transducers are available for sheath monitoring. Contact monitoring transducers for monitoring contacts can be made addressable for use in remote locations. All of these transducers can be combined in a system due to their addressability. Transducers are the backbone of a proactive maintenance and monitoring system. They should be highly accurate, have a fine resolution and high reliability. Other factors to consider are high noise rejection, stability, low power consumption and mean time before failure. 4.2. Data Acquisition Equipment To gather the information from the field devices , such as transducers, a data acquisition unit is installed in the C.O. This unit collects the information from all the sources and sends the results to the monitoring computer. This acquisition unit should send information at any time it receives an alarm condition from a field device. It should not wait until it is interrogated as the time from alarm condition to interrogation can be crucial for protection and quick response to repair the cause before an emergency arises. For this reason the data acquisition unit requires some intelligence in the form of a computer circuit. Since the acquisition unit is intelligent it can be programmed for self diagnosis and diagnostics of the field devices. This type of system is crucial to the performance and reliability of a predictive maintenance system. Without diagnostics, false alarms could be registered or real alarm conditions may not be detected due to monitoring equipment anomalies or failure. Data acquisition units should be small for small C.O.s and expandable for use in large C.O.s. Some units also have the capability to read older type resistive transducers by using optional circuit boards. All units should have built in capability to monitor contacts as discreet alarms by contact is the simplest way to monitor equipment and buildings. The best systems are adaptable and expandable to allow optically monitoring of fibre optic cables by adding additional equipment. Communications with an acquisition unit be should available through a modem and direct connect RS232 interface. This allows remote and local interrogation for diagnostics, leak locates and maintenance. The intelligence of the unit allows direct or remote programming and automatic programming from the monitoring computer. Upon alarm condition the acquisition unit must be able to call more than one number in case of first number failure or if monitoring is done from more than one site, such as a dispatch center and a supervisory or management center. 5. Software 5.1. Operator Interface To use a software monitoring program we must have an interface between the software and the operator using the program. Just as the keyboard of a computer is the interface between the operator and the computer, the software information on the computer screen is the interface to the monitoring system. Information flowing from the field to the screen must be easily understandable. This interface should be designed for the operator and not the software. When operators spend time deciphering software instructions or information on the screen they are not being utilized for the main function of monitoring. An interface must be easy to see, navigate, conceptualize and control. Many operators have little knowledge of modern computer systems and that is appropriate. The operator is there to ensure the monitoring system is correctly programmed and information for proactive maintenance is readily available. The computer and software are only tools. You do not have to know how to build a pressure gauge to be able to correctly use one. The software interface must provide familiarity and intuitive choices to the operator. Shape coding, color and menu design incorporating all the different aspects of computer capabilities and human aptitude into an interface will produce a usable product. Information and instructions on the computer screen are usually organized in one of two ways, syntactically or semantically. Syntactic knowledge is required by the operator to use alphanumeric based or non graphic systems. This knowledge is usually short term and learned by rote. It is frequently dependent on computer system devices and void of relationship to the actual field devices. To execute a command a certain function must be memorized such as pressing a combination of keys (i.e. "Ctrl-A"). The problems with such an interface are that it is system dependent and may differ on different systems, organization to make it meaningful to the operator is difficult, it fades quickly from memory when not regularly used and is difficult to learn. Semantic knowledge is required for use of graphical interfaces. A graphical user interface (GUI) is the concept behind windows based software as opposed to language based software. Concepts and relationships are semantic. This knowledge is easily learned and retained because it is task based. System users will understand what is on the screen just by pictorial representation. An example is that if a transducer is graphically represented on the screen instead of alphanumerically it easily recognizable. A picture is worth a thousand words. The interface should have a menu system for organization of information. This menu system should be hierarchical so that relevant information on different menus is easily conceptualized and the system is freely navigated. In this way even complicated plants can be quickly traversed by structuring the menus in levels according to their hierarchy. Color coding greatly improves an operator's ability to perceive conditions of the system. Colors should be restricted and used sparingly and consistently to avoid confusion. A change of color should be used with a change of state as it is easily recognizable and distinguishable from other attributes on the screen. If we have an information database of a plant we can graphically draw the plant on menus. For example, we could have a picture of a computer on the screen along with all the data acquisition units for a system overview menu. If a data acquisition unit was shown in red we would know we have an alarm at the C.O. where that acquisition unit is located. If we move to the next menu (by using a mouse and double clicking on the picture of the acquisition unit in red) we should have a second level menu that pictures all the cables and equipment monitored from that C.O. On this screen we should see one or more items colored red. If we move down one more level, by clicking on the cable that shows in red, our screen should show all the transducers, pneumatic routes and/or branches of the cable. At this level we should see the transducers in alarm colored red and their corresponding readings. From this level we should be able to use leak locating functions and access all data pertinent to each transducer. This example uses basic semantics and shows how we can graphically represent the plant and alarms. 5.2. Outside Plant Database The database that must be programmed into the software module must accurately represent the plant. Incorrect information will result in incorrect assumptions by the software and the operator. A good software module will have limits on inputs for devices and levels of programming. This will reduce operator input error due to typing errors. A library of pre-drawn graphics of the devices will make the software easy to program. All other inputs should be bound to specific rules of the devices themselves so that the software cannot be programmed with information that is not functional. A good software program will lead the operator through the steps of programming. Use of legends for input is desirable for error free and quick database programming. The other part of the database consists of the actual readings received from the field devices. These readings along with the programmed database information should be stored in an open database structure for manipulation by other software programs for management purposes. A good software program will allow adjustment of the amount of information that is to be stored in the database. The program should allow for simple backup of the database so operators will do frequent backups of current information. Backups are an integral aspect of a proactive maintenance program. 5.3. Flexibility Flexibility or adjusting of controls of the software program to allow for customization is desirable to a certain extent. If flexibility is built into the program, operators can customize aspects of the program to fit the actual needs of their companies individual requirements. Some aspects of a program must be flexible such as the size of the database and the frequency of acquisitions of field information. Communication parameters must be flexible enough to preclude an alarm being missed. The program should allow password flexibility for different user levels. Adjustable controls for alarm thresholds, alarm reporting, and units of measure are all customized by the operator according to the conditions and changes of the outside plant. 6. The Complete System 6.1. Alarms and Communication As previously stated an alarm (a measured change in status beyond a certain threshold) must immediately be communicated by the data acquisition unit which continuously monitors the cable plant. This requires a dedicated connection between the data acquisition unit and the monitoring computer. The most versatile way to establish this connection is by a modem and dedicated telephone number. The data acquisition unit will call the first number it has programmed in its memory. If it fails to connect with that number it dials the second number, which could be a backup computer, and failing another connection a third number, which could be a pager, and so on. So as not to have a multitude of numbers to keep trying, the data acquisition unit should call one to three numbers depending upon the monitoring setup. These numbers should be tried in sequence and in increasing time intervals and after no connection to the main first number after a set number of attempts an emergency fourth number would be dialed. Once the main monitoring computer has acquired the alarm from the data acquisition unit, processing of the information received is done. The software analyses and compares the information to the database information to ensure a correct condition of alarm exists. When this has been completed (less than one minute from the time the call was received) an alarm report can be generated. This alarm will be visible on the screen and can be routed to a printer. By accessing the graphical user interface an operator can quickly determine what device or devices are generating the alarm (showing in red on the plant graphic). This alarm should be made available to other supervisory computers automatically. The alarm can also be sent automatically to an alphanumeric pager. If a field technician receives the alarm he can immediately proceed to the C.O. that generated the alarm and begin repairs. Some companies presently do this and have eliminated the monitoring dispatcher. Other communications consist of measurement information acquisition, diagnostics and database transfer to the data acquisition units and supervisory computers. Transferring the database to the data acquisition unit precludes calling the unit constantly for updates and allows the unit to immediately report alarms as it will have acquired the threshold information Therefore measurement information acquisition can be automatically done once a day or when required. Self diagnostics are constantly being performed to ensure the equipment is working properly. A good data acquisition unit shall have a separate means of communicating diagnostic information other than to the monitoring computer. We do not want to employ the monitoring computer for diagnostics and testing. It must be available to monitor for alarms. Access to the data acquisition unit, by modem or RS232 interface, and communication to the software within should be available by using a test device or a simple communication program on another computer. 6.2. Leak Locating Leak locating formulas that are extremely accurate are not generally available because they are developed by the company that supplies the monitoring system and therefore are proprietary. A good monitoring system must utilize an accurate leak locating formula. Once the measurements of the field devices have been acquired, by either an alarm condition and transmission from the data acquisition unit in a C.O. or an acquisition from the monitoring computer leak locating can begin. The software should provide a graphical indication of the leak location and an alphanumeric display of the exact position of the leak. The information used to find the leak will consist of a set readings taken at a specific time. The graph will show a leak that is calculated from absolute information. If the software allows us to store a previous set of readings of the normal conditions and calculates the leak by utilizing the difference between this set of readings and the new readings we shall have a more accurate leak location. This is considered normalized leak locating. As the computer does the calculations very fast and the operator has located the leak quickly due to the ease of use of the graphical interface he can dispatch a field technician to the exact location of the problem in a minimal (5-10 minutes) amount of time. The repair will be attended to before there is a problem with the cable. If a field technician is paged by the monitoring system and he has a laptop computer with the leak locating part of the software therein, he could quickly locate the leak himself. After the repair he could also ascertain the condition of the plant with the software to view the results of his work. This is also useful for upgrading the plant for an efficient network. 6.3. Allocating for Proactive Maintenance To benefit from a successful proactive maintenance monitoring system proper allocation of resources is a requirement. First of all a plan must be developed for implementation. This plan must be adhered to for the success of the installation and program. The computer itself will not solve all the problems with a network, but rather it is a tool that can be used to prevent the problems from reoccurring. Purchasing a computer monitoring system does not change a maintenance procedure. It reduces the reactive maintenance required and enhances knowledge of the outside plant. We must plan so that the best use and value can be had from the tools we purchase. Planning the outside plant should be the first priority. How many transducers and where to install them must follow system engineering guidelines. The supplier of the monitoring system should be consulted at his stage. Training for managers and planning engineers on general installation and use of a system is recommended before installation begins. During this phase the technicians should be allocated to the program for installation and maintenance of the system. These technicians can then be trained by the supplier on installation techniques. Once the planning and initial training has been completed and the system components have been purchased the installation can begin according to the engineered system plan. The allocated technicians can begin with one C.O. which will become the field training site. Time must be allocated for training of the technicians and operators on computer monitoring at this time. When the first C.O. installation is partially complete (some transducers installed and working) the monitoring software can be installed and the Installation verified. After the operator has received preliminary training the database programming can commence according to the system plan. Once programmed the system can be used for monitoring while it is being expanded. The installation technicians can respond to the system even while installation is taking place. The benefits and savings of the system will begin to be realized. This becomes the pilot project and everything learned during this installation will make expanded installations easier and enhance the growth of the system The emphasis in allocation is on the training for the personnel assigned to the maintenance. When properly trained these same technicians an also do the installation. Untrained personnel doing improper installation will create problems. Education of personnel contributes to the overall success and acceptance of the proactive maintenance program. 7. Summary A good proactive maintenance program and monitoring system is a combination of the latest technology, educated personnel, and management support. Accurate field devices strategically located, combined with a reliable data acquisition unit comprise the basis of the system. User friendly state-of -the art software automatically processes the information received from the field and personnel respond in a timely manner. Achieving a very successful proactive maintenance program requires planning, training and a good monitoring system. Adequate implementation planning and cross training will ensure the highest use and best value of the monitoring system. Preeminent software can reduce the amount of training required with on- line help and a graphical interface, and will secure quick acceptance of the system. The savings after implementation of the system will come from many sources. They include improved personnel utilization, reduced overtime, improved plant performance, reduced downtime, increased plant reliability and reduced emergencies. The major savings come from reduced loss of revenue when the plant is shutdown (cable loss) and equipment replacement cost due to failure caused by ineffective maintenance. canadian puregas equipment limited
Back to Canadian Puregas home page | E-mail Canadian Puregas
90406