Trouble with the Performance Curve – BuildingIQ


BuildingIQ Named a 2013 Enterprise Smart Grid Leader by Groom Energy
July 18, 2013
BuildingIQ Appoints Michael Nark as Chief Executive Officer
September 19, 2013
Show all

Trouble with the Performance Curve

Pete Dickinson, CTO
July 11, 2013

Straight lines are a little boring. Curves are where the action is at. Particularly with large buildings and air conditioning. Nothing works in a straight line. If only it was as simple as asking for 10% more output resulted in 10% higher power demands! Unfortunately things just aren’t quite that simple. Whether at an individual component level or at a total system level almost every type of equipment exhibits what is typically called a ‘part load performance curve’ when comparing output work with input power. Fans, impellers, compressors and such all demonstrate certain ‘sweet spots’ where the efficiency is at it’s highest. Conversely, there are load points where efficiency suffers.

When individual components are combined into a systems level view the sum of the curves can take on very unique characteristics. Characteristics that encompass not just the particular components but the layout and general design of the building itself. This is a key reason why each building is unique. That uniqueness can be characterized by system level performance curves. These can tell us how much heating or cooling we may get for a given input power during a certain time of day or a certain weather condition.

Putting aside the complex issue of extracting these system wide performance characteristics for a building, let’s discuss the gold that’s available in those curves.

Where is the gold? In this case it’s located in two places:

1) Avoiding Diminishing Returns. For example: increasing the load on plant whereby a LOT more power is required to get a little more output. That’s not good. Let’s try and avoid this.

2) Extracting more output for comparatively little extra input. For example: there may be times where we can get a lot more output for only a little more power input. Getting more for just a little extra? That’s awesome. Let’s try and do more of that!

Great! Let’s assume we know the total system performance curve. Time to mine some gold.

Not so fast. Let me digress for a moment…

If a building simply runs to a time clock with basic control loops then that building simply ‘rides that curve’ throughout the day to give the required system output. The plant is the slave to the thermostat master. 5% more cooling required? 5% more cooling delivered. How much power did that require? Who knows? Who cares? Simple as that. The building is a slave to the curve. Gold is wasted. The plant might be ‘tuned’ to the highest efficiency but clearly there’s significant opportunity in presenting the right load at the right time.

Let’s re-prioritize this relationship.

We must change that paradigm if we want to extract opportunities and avoid penalties. To do this we must introduce a small amount of modulation to the system – a ‘degree of freedom’. A small amount of variation can give us enough opportunity to be smart with presenting an intelligent load to the plant. Where does this intelligence come from? That’s where we must use a forward looking optimization system which can look forward and know where to find the opportunities presented. Not just from the performance curve but from a range of other areas as well.

Consumption price, demand price, weather, economy cooling and many other operational complexities present further gold mining opportunities.


Don’t forget to follow me on Twitter: @Pete_BIQ
Have you read my last blog post?