In this episode, we'll go through a couple of scenarios and discuss Cat's patented paralleling failsafe mode. Feedback? Email us at powerbytes@cat.com
Email us: powerbytes@cat.com
LinkedIn: https://www.linkedin.com/showcase/cat-electric-power/
Facebook: https://www.facebook.com/Caterpillar.Electric.Power/
In this episode, we'll go through a couple of scenarios and discuss Cat's patented paralleling failsafe mode. Feedback? Email us at powerbytes@cat.com
Email us: powerbytes@cat.com
LinkedIn: https://www.linkedin.com/showcase/cat-electric-power/
Facebook: https://www.facebook.com/Caterpillar.Electric.Power/
Lou: Good Day and welcome everyone to Power Bytes! I am your host Lou Signorelli and Power Bytes is your destination Podcast for power generation topics. As always please know how much we appreciate you, our listeners. We hope you find our topics helpful and interesting. There are several ways for you to get in touch with the show. You can send us an email at powerbytes@cat.com, visit us at Cat Electric Power on Facebook or LinkedIn. Please leave us a 5-star rating wherever you listen to your favorite podcast; It really does help
Lou: Keeping with our April and May theme of advanced generator set controls, today's we'll focus on what happens in a paralleled Cat generator set System when we lose communications with one or more gensets. Cat's patented technology is called Failsafe Adaptive Load Sharing. Here to share some insights on the technology is Ted Snider.
Ted has been a service training instructor for Caterpillar Global Dealer Learning since 2010. His responsibilities include development and delivery of Cat Electric Power related content for web-based and instructor-led training. His primary focus is electric power controls training. Ted has directly trained about 250 Cat technicians on genset controls since 2016.
Thanks for joining us Ted!
Ted: Hi Lou.
Lou: Ted, many of our users have multiple generator sets in parallel. Clearly keeping them up and running is important: as is keeping them balanced even in the face of the most troubling of situations such as the loss of communications. For our discussion today, let's assume that our user has five (5) 2MW gensets running in parallel. In a moment we'll walk through an example of a failure, but first can you speak to how the gensets are connected for the purpose of communicating?
Ted: If the gensets are interconnected using the Multiple Genset Data Link (or MGDL), they will communicate over an ethernet cable using our proprietary protocol. These units will typically be connected together via a network. With the MGDL system, each genset controller maintains memory of all the units that should be present. When all units are accounted for, the system will manage paralleling over the network. The MGDL network allows the gensets to coordinate all the vital paralleling functions like dead bus arbitration and sharing of resistive and reactive load. If the load sense/load demand feature is used, this is also handled over the MGDL network to manage which units will stay on and which units will go off the bus.
Lou: Thanks Ted. We just wrapped up a podcast that aired in early April which included a discussion of load sense/load demand.
In our example I mentioned that we had five (5) 2MW gensets in parallel. To continue with that example, let's assume the average running load is 8MW. Can you describe the techniques used to keep them in parallel?
Ted: First, when a run signal is given, all the units will start. Dead bus arbitration over MGDL negotiates permission so the first unit to reach rated voltage and frequency closes to energize the bus. Next, all the remaining units will use their local sensing to synchronize to the live bus and close to it. As load is added to the system, the units will actively balance the kW and kvar load. Since all the units communicate their breaker status (connected/disconnected) and load factor on the MGDL network, all online units can balance load using their local fuel and excitation control. With 8MW on the bus, each unit would carry 1.6MW.
Lou: Thanks Ted. Now for the fun part. Let's assume that something happens and we lose communications to two (2) of the gensets. The average running load is still 8MW, how will the units respond?
Ted: Here’s how the units will respond:
• The three units that are still communicating will enter Failsafe Isochronous mode. The two units with lost communications will enter Failsafe Droop mode.
Let me expand on the failsafe droop vs failsafe isochronous modes….
In this case, the expected number of controllers is five (5). When communications are lost with the two units, the three remaining units see fewer than expected. However, since this group has a majority of the previously known units, they will Proceed as Normal and enter Failsafe Isochronous mode. The two other units expect to see 5 units, but only see 1 (or two if they are still communicating). Either way, they will “Proceed with Caution” and enter “Failsafe Droop” mode. All units will maintain their current load factor unless the system load changes.
Lou: Ted, can you take us through what happens when load Increases?
Sure, If the system load increases, this is where things get interesting the isochronous units will accept most of the load increase. The isochronous units will maintain rated voltage and frequency until they exceed about 80% load factor. As these isochronous units raise above 80% load factor, they will intentionally droop frequency (or voltage on the reactive side). This droop action will cause the droop units (proceed with caution units) to accept load according to their calculated droop curve.
This is a clear benefit to our system over simple droop/isochronous load balancing. In those systems, the isochronous units would need to be overloaded to decrease frequency. In Cat's failsafe mode, our isochronous units will intentionally droop before they reach overload. This allows our system to remain capable of supporting full system load capacity.
Lou: Ted, can you take us through what happens when load decreases?
Ted: If the load decreases, the isochronous units will drop most of the load decrease. The isochronous units will maintain rated voltage and frequency until they fall below about 20% load factor. As the isochronous units decrease below 20% load, they will droop up. This droop action will cause the droop units to decrease load according to their calculated droop curve.
Again, this is a benefit over simple droop/isochronous load balancing, because in those systems, you’d need to enter reverse power with the isochronous units to decrease load from the droop units.
I should also mention that the two non-visible units will also decrease their loads at low system load factors without pushing the isochronous units into reverse kW (or kvar) and causing a shutdown fault. This allows us to keep all our generators online all the way down to zero system load.
The reason this works is large part due to our patented technology. The droop action of the isochronous units (at high and low load factors), and the adaptive droop of the droop units.
Lou: If/when the load changes how often do they adjust their droop curves and is this an automatic function?
Ted: The droop curve for the droop units will update based on the frequency (or voltage) changes detected by the unit. The droop units will calculate their Load Increase or Load Decrease slopes whenever there is a frequency (or voltage) change. This can be due to a system load fluctuation, or due to an overall increase (above 80%) or decrease (below 20%) in system load.
Lou: Ted, now let's focus on the 3 units we can see. How will they respond to changes in load on the system?
Ted: The units that are still communicating will share their load proportionally across their ratings. When they are below 20% or above 80% load they will droop up or down together but still balance their load factors. Between 20 and 80 percent, they will maintain the bus frequency and voltage at nominal values.
Lou: Ted this is fascinating. And all this is controlled by the genset controller on the genset? Does each controller act independently? How are they talking and who is in charge?
Ted: In the Cat Genset Control system there is no one controller that takes charge. Any unit can be a proceed as normal (in isochronous mode) or a proceed with caution (in droop mode) unit depending on how the communication outage occurs. And it gets even better. If communications is restored, the units can go right back to proceed as normal and perform regular load sharing action without operator intervention.
From the demonstrations I’ve performed in my lab exercises with the Cat techs, it performs so well, some customers may not even realize they have an issue if they don’t look for the cause of the amber warning lamp on the genset controller or at their main monitoring hub.
Lou: Ted, what happens if we lose communications before a run signal has been given?
Ted: So if the units lose communication before they receive a run request, they’ll perform a failsafe dead-bus arbitration routine. The proceed as normal group will arbitrate as normal, with communications based permissions. The proceed with caution group will dead-bus arbitrate with a timer value according to their assigned unit number. This minimizes the possibility of a simultaneous dead bus closure.
In either case, the proceed as normal and proceed with caution units will exit dead-bus arbitration and synchronize if they detect a live bus with their local bus sensing.
Once they're on the bus, they will enter Failsafe Isochronous mode or Failsafe Droop mode like we have discussed.
Lou Summarize: Ted, I'm hearing a lot of great benefits for owners of Cat genests due to the patented Failsafe Adaptive Load Sharing. Let's see if I have them right
1) The Failsafe Adaptive Load Sharing technology assures loads are balanced across the "visible" paralleled gensets and Adaptive Droop Mode keeps the "invisible" gens online and protecting load
.2) the entire capability of the system remains available and the buss remains alive even when communications have been interrupted.
3) Because of the patented technology, we are also preserving generator set health by not forcing units into overload to accommodate certain balancing functions. And 4) Failsafe Adaptive Load Sharing also prevents any reverse power or Kvar which can lead to unit shutdowns.. Did I get those right?
Ted: Yes, and add two more.
1. We maintain rated voltage and frequency across the bulk of the load range, only drooping at very high and very low load factors.
2. We can get all units online without problems if communications are lost prior to receiving a run signal.
And all this happens without operator intervention.
There you have it folks. Thank you for joining us today on Power Bytes. For more information on this or any of our podcasts contact your local Cat dealer or visit Cat.com. If you’d like to suggest other topics for the program or have some feedback to share, please write us here at powerbytes@cat.com or visit Cat Electric Power on Facebook or LinkedIn. As always don't forget to subscribe to this podcast. Have a great day.