"This approach would support going beyond simple ON/OFF to analog control, so the problem is much more complicated," said Morgan. "It may also include expansion of the base and electrical energy objects to include the concept of demand, or average power over time, at the device level."
While idle state energy reduction offers an ON/OFF capability to reduce energy usage in devices not in use, phase 2b would target more of a real-time control loop. Then, the user might give the application a target to not exceed a specified amount of kilowatt usage. Based on energy awareness and the power management objects in its collection of devices, the application would decide on the control actions to take to satisfy the set points.
The setpoint could be established by a plant owner or process owner who doesn't want to set a new demand peak for the month, and, based on that new peak, pay for it for the next year. Or, potentially, the setpoint might come from a utility provider or energy service company because the grid is having a demand-response situation and they are requesting their customers to reduce power consumption. It's a puzzle piece that supports the overall picture.
Creating deep sleep in the devices proved to be a special challenge. The Ethernet/IP adaptation, for example, adopted Wake on LAN technology that is used in notebook computers. The approach is based on a specification developed by Intel and IBM, and documented in an AMD white paper. It does take specific Ethernet PHY and MAC hardware to implement deep sleep, but fortunately, the hardware is found in commercially available Ethernet products.
Impact on energy management applications
"The trend we see with these developments is an ability to make energy information in the industrial setting more granular. The typical way to measure energy today in the factory is by using power meters at different locations on the electrical distribution system within plants," Michael Crowley, product marketing manager for Schneider Electric, and a member of the ODVA working group, told us, going on to say:
But if the energy object is widely implemented by control vendors and machine OEMs, it will be much easier for the industrial user to understand energy usage at a much more granular level. Companies should be able to monitor energy consumption in terms of an individual process rather than strictly the electrical distribution system, which is the way we do it now with the use of meters.
Crowley says that if the energy awareness capabilities are implemented inherently in a machine, it is less expensive and more convenient. There is no need to retrofit equipment to get that additional information, or spend extra money for specific metering since those capabilities are built into the equipment itself. End users want to understand where they can save energy and, to do that, it's important to understand where energy is used within the process.
Using the Energy Object as part of the network specification makes it easier for users to integrate energy information into their energy management systems. Today, vendors' devices talk well with their own products, but integrating third-party energy information can be expensive.
If a standardized way to describe energy usage and exchange energy information between vendors emerges, it will potentially become much less expensive to understand how energy is being used. At the energy analytics level, this process is dependent on what energy management system a plant is using. But the end result of this development should produce good information presented by a good reporting tool, so the user can both understand it and manage energy usage more effectively.