Conserve It R&D Engineering Manager Michael Berger addresses some fundamental concepts in better understanding the purpose and challenges of a controls strategy for chiller plant efficiency.
The first concept I am going to address is about sensors and meters, and why their accuracy is essential for what we are trying to achieve.
The first term I should probably define is Energy Efficiency. The term has taken an all encompassing meaning in the past few decades of being the intent of reducing energy use of a component or system, but it is also commonly used to describe the physical number that quantifies how much energy the system uses to deliver a unit of useful output.
This number is calculated as the ratio of the useful output by the input, in terms of energy. For instance, for a machine or a system that produces cooling:
the useful energy output is the cooling produced that is supplied to the field
the energy input is the electric power that is being used by the system (the whole plant for instance) or the component (the chiller for instance)
This Energy Efficiency number allows to compare different equipment, system designs or control strategies on an equal footing in terms of energy usage.
For chilled water plants, this number is very sensitive to even very small errors in sensors. In particular in chilled water temperature sensors. This is because the cooling produced (sometimes referred to as "cooling load") is calculated by multiplying the difference of temperature between the return and supply chilled water line with the chilled water flow and some physical constant.
It is fairly common for a chilled water plant to be designed to operate at full-load at a chilled water temperature difference of around 5Δ°C, and because the cooling produced is directly proportional to this temperature difference, an error in readings that may appear small at first glance like for instance 1°C, would result in an error in cooling produced and efficiency of 20%, which is very significant.
To give some perspective, on a typical site with no other retrofit other than installing PlantPRO, the energy savings that are being touted are often around 10%.
As a result, in Singapore, arguably a pioneer and front-runner on the subject of regulations for chilled water plant efficiency, the building code that all new construction and retrofits must follow now prescribes temperature sensor uncertainty of +/- 0.05°C or less for chilled water, which is extremely stringent (and in my opinion justified), but which is not at all common around the globe and in Australia as of today.
Measurements accuracy is not only important for calculating the energy efficiency, but is also essential to good controls. A control strategy is only as good as its inputs, and poor data will result in poor controls.
The questions that ensues is: What can we do about it?
Well, there are limitations to what a fully automated system can do in these conditions. Although not full proof, there are ways (albeit somewhat limited) for an automated system to assess whether there might be sensors errors:
The Heat Balance of a chiller, which can be simplistically defined as "the energy input must equal the energy output" can be calculated if the appropriate sensors are present, and if there is a discrepancy it indicates that there is at least one sensor error somewhere.
Comparing certain sensors can highlight incoherence and indicate sensor error. For instance, if the chiller supply temperature and the plant supply temperature are very different, it may indicate that at least one of the sensors is reading incorrectly.
When the above issues are flagged, the software can automatically adjust some things (like disable some diagnostics that rely on the targeted sensors, discard data that would otherwise be used to calculate the machine performance, etc.), but the issue must often be corrected by operator manual intervention, who, upon being notified by the system, would have to troubleshoot and possibly calibrate or replace sensors.
Issues become opportunities
In practice, often times issues that are being reported by operators as originating from PlantPRO are actually due to data errors, be it inaccurate sensors or flimsy communication.
However, these challenges can be flipped upside down to be a great opportunity for PlantPRO. Our partners can pair PlantPRO with an on-going Maintenance contract, and act upon notifications from PlantPRO of possible incorrect sensors values, to fix these sensor issues early. This allows to save energy and money to their client by ensuring that the system does not operate for weeks or months at a sub-optimal efficiency due to unnoticed sensor errors misleading the controls.
In any case, until such times when we figure out how to fully automate fixing sensor issues, these challenges are and will be part and parcel of every PlantPRO site along its lifetime. And it is important to understand how they affect the ability of the solution to produce its expected results, as well as to consider any potential improvements that would help attenuate these issues.
As the R&D Engineering Manager at Conserve It, Michael leads the R&D efforts to develop innovative chiller plant optimisation and advanced controls solutions.
Having graduated with a Master of Engineering in France and with experiences in a leading research centre and in ESD consultancy in Singapore prior to joining Conserve It in 2014, he relies on years of experience in the fields of Energy Efficiency, Chiller Plant Optimisation and Energy Modeling.