Automating the After Hours HVAC and AC Extensions process – Part 1
A popular service that you can offer to your tenants or staff is the ability to let them keep the air conditioning running after the close of normal office hours.
Typically most buildings run their cooling on an automated schedule, for various reasons, the most important one being energy savings.
However, people won’t work on automated building schedules, and you occasionally find the need to keep the comfort levels up even after normal office hours to accommodate staff burning the midnight oil.
The primitive way this is done is that someone tells someone they will be working late so please make sure the air conditioning is kept on.
The message eventually gets thrown into a bottle, into the sea and picked up by the building’s facility team which will then make the required arrangements.
But things are smarter now. Modern technology can make the whole process seamless and even invisible.
Once you book your meeting, ACs are automatically taken care of.
It’s surprisingly complex.
In Part 1, we’re going to look at the complexities of running a centralized heating, ventilation and air conditioning systems on a schedule and what problems you often come across.
In Part 2, we’re going to look at different strategies for making the AC extension process seamless.
Understanding Cooling Systems
Unless your workplace is small and fitted with individual air conditioning units, your building very likely has a centralized heating and cooling system.
This involves the supply of cold water from chillers, heat exchange, circulation of heat, ventilations, refrigerants and a lot more.
Here’s a quick introduction to the various components involved.
HVAC(Heating, Ventilation, and Air Conditioning)
The entire heating, cooling and ventilation system that delivers comfort to your workplace is called an HVAC system.
The basic principle is around transferring heat from air to chilled water and then circulating the cooled air.
So there are two classes of equipment in an HVAC system: The water side and the air side.
Water Side Equipment
These consist of among other bits, chillers, cooling towers, chilled water pumps, heat exchanges. The equipment don’t really behave as individual components snapped together. Instead they are custom-designed and architected and put together to meet the specification, requirements and shape of the building they are working for.
Chillers are probably the most important component of the water-side system. They are responsible for supplying chilled water. These are typically placed in the basement of a building, or even outside, to reduce noise. Heat exchangers are generally installed outside, placed to avoid disrupting neighboring buildings. If cooling towers are used, they are usually installed on the top of buildings (unless the roof is ear-marked for solar panel rental).
Typically there will be a group controller orchestrating the functioning of these systems but you do also get systems with individual controllers in them.
It’s also a growing trend for buildings to not have chillers or cooling towers at all, and instead use a ‘District Cooling Plant’ – a central supply of chilled water on a commercial arrangement to multiple buildings.
Air Side Equipment
These are the ventilation systems, the (Variable air volume)VAVs and (Constant air volume)CAVs and air diffusers.
At the boundary of the water and air side, you get Air Handling Units and Fan Coil Units.
AHUs
The Air Handling Units are responsible for circulation of air. They generally take cooled air (via heat exchange with the chillers) and send it into the ducts for distribution through the building. They can also mix in some outside air to ensure carbon dioxide level within the building is maintained at a safe lavel. In regions where with multiple seasons, bringing in cold / warm air from outside can increase the efficiency of the cooling system.
(During cold seasons, heating can increase humidity, so sometimes, paradoxically, cool external air is mixed to reduce humidity)
Humidity of the environment directly affects comfort, therefore, it has to be controlled. Sometimes humidity has to be reduced, in which case temperature of the cooling coil of the head exchanger is reduced below dew point and then supply air is reheated via a heating coil to arrive at the desired temperature. When humidity has to be increased, humidifiers are employed before supplying air to the environment
AHUs typically don’t have a controller attached to it to program its behaviour. The BMS supplier will often take that responsibility of installing a controller to monitor and control the AHU equipment.
FCUs
The Fan Coil Units are similar to AHUs but are simpler and more self-contained. They circulate air within a self-contained area and do not use ducts to move air around and do not mix air from the outside.
Automation of Cooling Systems
Chillers are complicated equipment and they have a complex start up and shutdown procedure
They are typically programmed by the BMS vendor to run the start-up sequence in advance of the first working hour and typically run their shutdown sequence as soon as the working day ends.
This sounds simple enough but there are plenty of hidden complications in this.
First we need to understand what actually goes on when the chiller starts up and shuts down.
The Shutdown Process
In order to understand complexity in start-up sequence, we need to talk about the shut-down sequence first.
When a chiller is shut-down, it’s not as simple as turning off a switch.
There are sub-processes that have to keep going for a while.
For example, one process would be the removal of refrigerant from oil by boiling it off.
Once the chiller is operationally down, some of the valves are closed for safety of the equipment and the people around it.
The Start-up Process
Chillers are not an independent equipment that can be turned on and off arbitrarily. They need to work in sync with cooling towers and heat exchangers.
In most of the buildings, these equipment together with connecting pipes and valves are custom designed based on the nature and the configuration of the building.
It isn’t easy to automate the start-up sequence without some manual steps.
The first assumption of the start-up sequence is that the shut-down sequence was performed correctly. Various sub-tasks have to be performed for the chiller to be ready to start, such as running oil heaters to make sure any leftover refrigerant is separated from oil.
Automating start-up and shut-down.
As such, it is not operationally possible to turn a chiller on or off arbitrarily because both shutdown and start-up sequences take a certain time to work.
There are two reasons why chillers need to be shutdown completely:
- Prevent energy being wasted from running chillers with no load requirement from the building
- Remove the requirement for a qualifying technician to be on duty to watch over chiller operations.
Unless there is a significant gap between shutdown and start-up times, it doesn’t always make sense to shut down chillers.
There are many buildings that run at least one chiller at all times and only start additional chillers as load rises.
Automation Strategies
In situations where the chiller is always running, it is quite straightforward to automate the air-conditioning by turning on/off AHUs and Air Diffusers as necessary.
In this situation you would create a mapping between sets of A/C equipment and spaces that users occupy – because users do not know about or care about individual A/C units that are serving their spaces.
When air-conditioning is not required for any of the dependent spaces, the set of equipment can be turned off.
However, in case of chiller operations, a logical process has to be implemented to determine ‘shut-down time’ and ‘next start-up time’. If there is a significant gap between them, it would make sense to shut the chiller down.
If the chiller start and shutdown sequences are already programmed and triggered using a scheduler, the most convenient control strategy would be to override the schedule, such that rest will be taken care of by the existing automation logic.
Air-conditioning equipment such as chillers and cooling towers do not come with their own controllers to provide digital control. Often, they are ‘hard wired’ to custom controllers provided as part of BMS. Since these equipment are not physically closer to each other (one in the basement while other on the rooftop) they are controlled by different controllers. When it comes to automation, in some situations, a single group controller is designated to handle the logic by communicating with individual controllers to execute commands.
In other situations, sequence logic could be distributed across individual controllers. For example, scheduled starting sequences may be triggered by independent schedules in each controller.
In such situations, great care should be exercised when ‘overriding’ these schedulers because failure to update one could be catastrophic.
This can be further complicated when the building has multiple chillers where each chiller works on specific days on the calendar (such as odd and even days) and the other acts as backup to handle high-load situations. Now, we also need to figure out which chiller is under operations on that particular day before overriding its schedules.
It is very common to use ‘bacnet schedule’ points to trigger start-up and shutdown sequences. Before attempting to override bacnet schedules, there are few exceptions to be noted.
First, recurring schedules are made using ‘day of the week’. This allows operators to schedule weekends to be different from weekdays. So, when overriding the schedule, it will not just affect next start-time or shutdown-time but also future operations of that day of the week. As such, it is very important to ‘restore’ the schedule once the day has passed. Alternatively, a process should be defined to override schedule for the next day on each day.
Secondly, there are exception schedules that temporarily overrides the recurrent schedule. These exception schedules will take a full calendar date and time. These are often used for forcing systems to behave differently on holidays. This could be another strategy that can be used to enable chiller extensions without modifying the recurrent schedules. An exception schedule can be placed to control the next operation.
However, this makes it difficult to ‘modify’ schedules (say, a tenant asked for a two hour extension but later asked for another additional hour). In such cases, previously placed exception schedules have to be modified instead of placing a new one.
