|
Remotely monitoring and controlling onsite power systems can improve their reliability and effectiveness.
Onsite power generation is often considered to be made up of three sub-markets: emergency standby power, rental (also known as mobile) power and prime (or continuous) power. The application of remote monitoring and remote control for these three market components requires distinct strategies. No single wireless technology is a best fit for all three. Likewise, customer data needs are significantly different across the categories.
Fundamental to effective remote monitoring is the concept that the remote monitor itself must autonomously detect changing operating conditions and communicate these changes to a common database where data is stored and served to a set of intended users. The old model (represented by a user dialing up a machine and viewing current conditions) is simply inadequate compared with today’s technology. Currently, third-generation cellular data methods allow continuous, real-time transport of operational data without the need for any customer infrastructure such as phone lines.
A design point in which all machine data is stored by a central, web-accessible database is essential to broad utilization of remote monitoring. This approach allows multiple users simultaneous access to the machines (and all of the archived data from the machines). It also enables a process in which every operation of a genset throughout its service lifetime can be presented and archived.
The standby power application is characterized by long periods of waiting for a utility power failure, interspersed with weekly automatic engine running exercises and occasional manual engine tests. The weekly exercises are critical to maintaining good engine operating conditions. The exercises also provide a key opportunity for remote monitoring. Specifically, the minimum requirement for monitoring standby power is to confirm the weekly exercises. From a statistical perspective, if we already know that the genset has operated successfully for 50 consecutive weeks, we will have a high confidence that the genset will operate successfully the next week. It is this high confidence level that we seek to achieve with remote monitoring.
The weekly exercise detection is a critical first step. But should there be a component failure between exercises, the first indication would be a missed exercise or a shutdown fault at the time of the next engine start. It would be entirely possible that the fault condition could exist for a week without monitoring other conditions. The most common causes of “fail to start” events are low fuel, low battery, failed block heater and being in a not-in-automatic condition. (The transfer switch and timer cannot start the engine unless it is in automatic)
Notice that all of these conditions are caused by consumables with predictable lifetime or user error. They are not due to brand or engine quality and will occur on all brands. However, they are all conditions that are completely detectable with basic remote monitoring. What’s more, the time between the start of the condition and its correction can be reduced from a week to mere minutes, essentially eliminating over 90 percent of all starting failures and further raising customer confidence levels. In general across all brands the fail-to-start rate is on the order of 1 in 1,000. With even the most basic monitoring eliminating 90 percent of start failures, the fail-to-start rate can be reduced to around 1 in 10,000. From the genset brand perspective, this can alternately be stated that 9,999 successful starts are expected for every one failed start.
Most standby power monitoring can be handled via the basic monitoring method known as “report by exception”, in which changes of condition on the engine cause the wireless transmission of the machine’s status. This monitoring method was the standard in first-generation cellular data products where data packets were small and expensive. The method provides instantaneous monitoring of the machine’s status, but not variable analog engine values such as temperatures and pressures. The history of the deliveries of status information becomes the activity archive for gensets monitored in this way.
A typical user web display, showing units that are currently running or in a faulted condition, has tables that expand and contract to accommodate the units in these respective conditions. Changes of condition depicted on the web display are accompanied by email messages to designated recipients’ desktop computers and mobile phones.
Table 1 is an example of a simple activity listing for an installed unit. The most recent status is at the top with progressively older status conditions lower on the listing.
The rental power / mobile power application is quite different from the standby power version. On average, gensets used in this niche are small diesel machines in the range of 50 kW to 250kW mounted on trailers and used in construction applications. Contrary to the standby application, these machines run regularly and commonly for eight or more hours, corresponding to construction crews’ workday schedules. The machines are typically operated manually and generally their operation is not critical in loss avoidance (unlike the case in standby power). While shutdown faults may cause inconvenience, the shutdown was likely caused by the work crew misusing the machine or ignoring fuel levels.
A significant requirement of the rental power remote monitor is that it must not drain the genset battery, but must operate in an always-on condition. The 100 mA power consumption of a cellular data unit used in standby applications is simply too high for these rental power applications. Current generation satellite methods can reduce the average power draw from the genset to less than 1 mA and can even operate entirely using energy drawn from a four-inch-square solar cell.
While the genset service organization certainly wants to know if its equipment is having shutdown faults, it may not want to be notified of every start-and-stop event. It primarily is interested in daily utilization reporting, total engine run time for service scheduling and in the machine location for logistical planning. Additional anti-theft features such as tamper alarming and geofencing (in which an alarm is triggered when the machine is moved from a defined GPS location) as well as low fuel alarms (triggering fuel resupply) are expected in today’s monitoring systems. Since rental equipment is expected to operate in arbitrarily remote locations, satellite connectivity is also demanded. Satellite connectivity also is important to support disaster response, where cellular connectivity may be at risk. Such was the case in the 2005 Florida and Gulf Coast hurricane season.
The prime power application has still a third distinct operating environment. While we often think of prime power in terms of generation in the jungle, it also includes the growing landfill gas generation application and certain utility peaking applications. Compared to the first two categories, this application demands the most engine performance data. Current generation customers expect to see that data delivered in a continuous stream. Whether it is engine temperature, manifold pressure, output phase values or power factor, the application is critical and users expect a high level of performance data.
In the standby model, detecting changes in machine conditions was adequate and appropriate. The likelihood of those machines being started at a given moment due to a utility failure is low, and less need exists for engine data from an engine that typically is not running. However, in the high performance peaking and demand response applications, just the opposite is true; gensets configured for utility or prime power are commonly run for long periods of time. These machines are usually equipped with engine controls rich in performance data. Across the major brands, the Modbus data protocol has become an informal standard. The Modbus connectivity method is an excellent bridge between simplicity and capability and is standard in the OmniMetrix real-time monitoring process.
In the real-time monitoring application space, the installed field device continually scans a list of engine parameters (Modbus registers). The list is scanned roughly once every second. As part of the scanning process each returned parameter value is compared to upper and lower threshold values locally stored in the remote field device. If the monitor detects a change in local conditions, it instantaneously pushes the entire set of engine analog parameters and digital status to the centralized database. Otherwise, on a regular basis (around every five minutes) the monitor pushes the entire set of scanned parameter data to the database. This enables the collection of key performance data into an ongoing archive of genset operation, available via web browser interface any time and any place. This method of continuous transport of engine data from the genset to the web database removes the need for the older method of a user dialing up a genset and testing it online. Such “one man, one machine” connectivity is now replaced with “unlimited users, unlimited machines” connectivity, since one user’s access to a machine does not block another user.
The use of continuous, real-time monitoring reveals surprising detail into remote genset operation. Such detail is shown in the engine temperature plot (Figure 1). The sawtooth data shape shows the normal block temperature variations due to the block heater operation. The large hump in the middle shows the engine temperature rising, holding and falling as a result of the genset running. Figure 2 shows the same machine at the time of a block heater failure.
The earliest (first generation) cellular data channels were capable of moving only small packets of data on a slow schedule. While somewhat limited, this technology was revolutionary in 1997 when it first came into use. This technology has served the report-by-exception monitoring method well, and has been shown to be dependable and geographically ubiquitous.
The current (third generation) cellular data process typically allows IP addressed connectivity between the field device and the web database server. This technology enables the transport of large amounts of data at high rates of speed. This method has enabled development of the real time monitoring technique.
For challenging geographic installations, satellite connectivity typically is used. The most affordable technologies are more capable than first generation cellular data, but still have low data rates. This keeps those technologies in the report-by-exception application space. A higher performance satellite connectivity also exists, allowing the implementation of real-time applications with completely global coverage. Such pole-to-pole coverage is important to multi-national equipment providers.
ISO Demand Response Programs: Current generation remote monitoring and control methods offer an excellent match for ISO demand response (DR) programs. Where such programs were initially designed around person-to-person phone calls scheduling demand reduction engine operation for the next day, according to ISO demand management plans, current generation remote control methods can enable 15-second command - response speed.
The real-time monitoring approach provides essentially instantaneous command, control and reporting capability to an arbitrary number of geographically concentrated or distributed gensets. The power produced by the numerous gensets is continually reported and summarized for utility consumption. Simultaneously, service-oriented engine data is continually displayed for the appropriate service companies, wherever they may be.
The process of invoking a demand reduction is as simple as clicking a web button. The ability to automate the dispatch of the demand response call to achieve a target demand reduction is fundamental to the process. From a pool of 150 MW of available gensets, 100 MW can readily be commanded in a preferred order with the desired demand reduction observable through the monitoring system. Should the reduction drop be due to a genset fault or simply due to variations in plant loads, additional generation may be dispatched automatically.
Automated Demand Reduction to Avoid Distribution Instability: When a distribution system sources its power over long transmission lines, it is possible to experience instabilities due to inadequate local reactive power or power factor correction capability. In particular, when a system becomes unstable, voltage dips occur and many heavily inductive motor loads disconnect from the grid to protect against poor power quality. When large numbers of machines move on and off grid in such circumstances, their startup current requirements exacerbate the instability, potentially cascading into a full power outage such as was seen in the August 2003 Northeast blackout.
Part of the benefit of a standby generator monitoring system is the ability to monitor and detect such distribution system instability in voltage, frequency and power factor. Under such situations, the remote monitor can cause a standby genset to start and pick up loads which otherwise would be attached to the grid. The monitor can instantly indicate that the process is underway and can deliver continuous measurements of the amount of power being provided by its attached genset. Note that this applies to standby equipment operating off-grid, not in parallel to the grid. The intent is to remove loads from the grid, not add power to it.
Whether autonomously activated by the sensed instability or under the command of the utility, it is possible to remove hundreds of megawatts from a distribution system within seconds. Such response can free up a volume of reactive power that can restabilize the system and avoid cascading blackouts. This approach can be achieved entirely from customer-owned gensets and with minimal utility investment.
Current generation wireless remote monitoring of distributed onsite power systems provides the ability to notify equipment owners and service organizations of successful operations and equipment failure events, as well as to remotely command that equipment. Incorporating GPS functionality with satellite connectivity makes it possible to track, monitor and control equipment anywhere. And finally, the ability to extract engine data via Modbus interchange allows a monitoring design point in which every operation of the machine throughout its service lifetime is recorded and archived. This capability closes the loop between the genset manufacturer, the machine owner and the service organization, providing optimum installed reliability and performance to the end user.
Harold Jarrett is president of OmniMetrix LLC, www.omnimetrix.net, based in Norcross, GA.
Power Engineering February, 2007
Author(s) :
Harold M. Jarrett
| 
Articles 
| View all PennWell sites | View all PennWell events
Add RSS Feed
