What does BMS tuning mean?

Tuning means different things to different people

During the final stages of a new construction project, tuning is an excuse to move defects that may hold up practical completion into the next 12 months of Defects Liability Period. E.g. the chillers aren't staging up very well; everything starts and stops okay but it's just not working very well. How many times have you heard: "this is a tuning function, we can tune this in the next 12 months"? Excuse my sarcasm, but I had to get that one out of the way, and to be honest, we have all played that card. So, in some cases tuning is a get out of jail free card. 

Lifecycle Controls Bryce Anderson independent BMS consultant


Most people believe that tuning is the adjustment of PID (proportional, integral and derivative) settings in control loops, resulting in stable control of the outputs (valves, dampers and VSD's) to maintain a process variable (temperature, pressure and volume) to a set point. This, of course, is tuning; and in the past this is all that tuning was. However, in my opinion this isn't anything special; it should be something that just happens and not something clients pay for every year in their annual maintenance contracts. Technically, once a PID loop is set up, it shouldn't need re-tuning again. 

Modern day tuning

In my opinion, tuning is staring at the screen for hours, thinking about each system as a whole  and not just a localised PID loop. Tuning is analysing how systems interact with each other and how small changes to any one system affect other interconnected systems, and the resulting overall effect on control, comfort, reliability, and ultimately, energy efficiency. For example: we all know that reducing the condenser water temperature will improve the efficiency of the chillers. We know that reducing the condenser water temperature causes an increase in cooling tower fan speed and therefore, power consumption. Chiller manufactures will usually say that the additional power used by the cooling tower fans is still worth the efficiency gained by the chillers. A chiller manufacturer once said to me: :"just run the cooling tower fans 100% all the time, we will still be better off".

Tuning starts with prep work and a plan, because tuning isn't a tech sitting on a drum of cable in a plant room with his laptop plugged into a controller, like it was 20 years ago. Firstly, we build new software points to monitor COP (Coefficient of Performance) for each chiller, all chillers, and the whole chilled water system. COP is the relationship between the electrical power (kWe) drawn by the chiller and the thermal power (kWr) generated by the chiller. As the COP increases, so does the efficiency of the chiller. 

For Example:

Lets say we have two cooling towers delivering a condenser water temperature of 21°C to two chillers. As we override the condenser water temperature set point down to 20°C, we will see each chiller's COP increase (more efficient), the cooling tower fan speed increase (using more power), and the total chilled water system COP increase. Great result, we proved that using more cooling tower fan power resulted in an overall system improvement. Now, as we continue to reduce the cooling tower condenser water temperature set point, we may notice that each chiller's COP continues to increase, but that the overall chilled water system COP starts to decrease. We have now determined at what point reducing the condenser water temperature, although improving chiller efficiency, has resulted in the total chilled water system to be less efficient. Not all chillers and cooling towers are the same; we can't apply the same rules for every system. 

This is tuning! 

Next, set up trends on each chiller's calculated COP, the cooling tower fan speed and the overall chilled water system COP. Group all those individual trends into a single trend and bind it to the chiller graphic (not in some folder buried in the server somewhere). Now, you have something to analyse. Click on the trend group from the chiller graphic and you can see the history and the relationship between chiller efficiency and whole system efficiency.          


COP= kWr / kWe

kWr = (Return temperature (°C) - Supply temperature (°C)) x Flow (l/s) x 4.181

Total chilled water kWe is the kW's from the chillers + kW's from pumps + kW's from the fans.

If you don't have a thermal meter on the supply chilled water to the field, you can calculate the total chilled water kWr from the sum of each running chillers kWr.

Some people use the kWr from the chiller High Level Interface. I prefer to use the differential temperature and flow (you can calculate l/s from the evaporator differential pressure transmitter if you don't have a dedicated flow meter). I believe that not all chiller controllers calculate kWr properly, as a couple of manufacturers have admitted when challenged.   







Bryce Anderson