MENU
Question: Ok, . . our critical load requirements combined exceed a single inverter capacity, can you explain alternative architectures?
Answer: Diversify the application by using smaller load bundles and deploy inverters with either INTERLACED or DISTRIBUTED architectures. These Alternative Architectures naturally build fault-tolerance and resilience into your application network.
Application Engineer’s Discussion:
This is a typical application question we field on a regular basis when designers telephone the office for product coffee chats. To draw a picture for you using automobile analogies, think of combining all of the family into a large transport vehicle for a trip to town. Sure, it’s just like the black and white television series with grandma’s rocker tied to the back bumper. Now, secure that vision in your mind. This model is combining all the critical loads on one inverter platform and heading over the hill. (My apologies to those not familiar with Buddy Ebsen aka: Uncle Jed and family…web surf the show)
Now, think about that 5 passenger sedan, of any brand suiting your fancy with well-equipped options, and visualize the reasonable seating arrangement plus the ability to separate those family members that don’t travel well together. You might even choose run-flat tires and a computer with SNMP trap technology. Everyone in the family is still going to town, just supported by two separate vehicles. It is recommended to place half the network load on one unit and the remaining half on a second unit. A visual along the lines of splitting parent and child processes across two inverters.
Sometimes we hear, “Well, that’s going to cost more.” That is a possibility or maybe not. A valid analysis comparing the initial investment and loss potential should be performed. If anything failed, will it cost more to be:
A. 100% out-of-service with your critical load (reminds me of a recent fiber cut to the office)
or
B. operational with a portion of your network and not totally in-the-dark?
Only the engineering designer of the critical network can weigh that value proposition and write an equipment design proposal to match the applicable network operational uptime goals.
Think about the logical grouping of the critical loads. We suggest grouping network elements together that help the application survive verses placing all the choke points or like equipment together on one device. Many customers even source the inverter dc voltage from multiple power systems with batteries, when possible. Full fault-tolerant diversity can be designed as the ac bypass circuits are pulled from two grid service panels. Why not tap separate bypass circuits, if the installation building has multiple ac grid legs.
A Summary Thumbnail:
1) If the load is so big that a transport truck is needed, then work it into the system and build the fault-tolerance with a bypass plan.
2) If the “combined” load is big because the collection of smaller equipment was added to a spreadsheet, then it will be well worth the analysis and planning to separate network elements into bundles.
Follow these links to see a graphic view of our coffee chat concepts.
INTERLACED or DISTRIBUTED
At the end of the day, it’s your network/application and we want to help you visualize a couple of alternative architectures for getting to town with the reliability and resilience required in critical power support using a Majorsine Power Inverter.
