Cutting energy consumption
If it works in the tropics, it can work anywhere
By Joel Shaprio and Thirumalaichelvam Subramaniam
While engineers specialize in specific manufacturing processes, few seem to focus on the overall efficiency of the plant, particularly in the use of energy.
Whether it is a car plant in Detroit or a calcium carbonate plant in Malaysia, being smart about how a manufacturer uses energy is such an important factor for any company these days.
There is now an approach that can improve the quality and energy efficiency of commercial air conditioning systems in the tropics, including those in office buildings, factories, and hospitals. In various applications, users were able to save up to 30% of their air conditioning energy costs in tropical countries where cooling costs typically consume 45% to 60% of building energy expenses.
The idea of cutting building energy cots really gained some momentum after going over the design and operation of a calcium carbonate processing plant in Malaysia back in 1989. This plant was one of the first to implement roller mill technology, a technique used to reduce the size of certain materials.
After the engineers built a model of how the entire plant operated end-to-end, the team quickly understood the operations were not economically feasible. The energy costs were too high, even in Malaysia where energy costs were largely subsidized by the government at the time.
After building that model of the plant’s operations, the engineers began taking raw data readings and mechanical observations and entered them into a database to study the characteristics of the inputs and outputs and model the plant operations. Through that modeling process, the plant’s energy bill went down by 20%, making it feasible to operate.
After this first success, they were able to work with a number of multinational manufacturing companies, continuing to model their plants and study and refine their energy consumption. This was the beginning of an energy consulting business.
By 1995, working with an American company to build hospitals in Malaysia, it was easy to realize the water cooling systems they installed in the hospitals were not well-tuned for the tropical climate because the temperature fluctuates to hot during the day and cool at night.
To cool a large area, commercial and industrial centralized air conditioning systems, like the ones used in Malaysian hospitals, traditionally consist of multiple machines known as chillers. These machines control air temperature by removing heat from water through vapor compression or an absorption-refrigeration cycle. Each air conditioning plant typically contains four to six large chillers, and each chiller operates independently to produce a return water temperature at a particular setpoint.
The typical approach to cooling a building is to determine how much energy you need to bring a particular building to a desired temperature and then set each chiller to produce chilled water at the same setpoint. The calculations usually lean toward the worse-case condition, the one or two hottest days of the year, which leads to inefficient energy use and hard-to-control building temperatures.
While those calculations work in a general sense, the results could be better if there was a way to manage the temperature of the chill water in real time, taking into account the spikes in the external temperature and the time lags, and to optimize loads across different chillers.
That began a search for technology to implement real-time data acquisition and temperature management across multiple chillers.
The solution was able to use a programmable automation controller and software to handle real-time acquisition of data coming from multiple sources in parallel, analyze those signals in real time, and send signals to control various motion controllers. The next step was to use the same approach to manage chill water temperature in real time.
The initial job was to run the hospitals’ cooling systems, but energy management was the ultimate goal. That meant hooking up sensors, putting them indoors and outdoors and looking at elements such as temperature, humidity, water flow, and electricity consumption.
The team started to think about the program they needed to analyze the data and control the heat transfer. They wanted to capture the outside air temperature and the water parameters in real time and conduct a dynamic analysis of the chilled water compared to the desired building temperature. The goal next was to synchronize the operation of the chillers, which may have come from different manufacturers, and their associated components and operate them as one thermal system.
A bigger team
The next step was to spread the wealth. With support from investors and a $1 million grant from the Malaysian government, four engineers were hired to begin training on the programming language to model the envisioned system.
Originally, the requirements for the basic program were written on calculators because not very much was known about software programming. But after learning how to program using new software tools, the team spent about a year programming all of the modules and then spent another year refining and optimizing them. The team has since added modules and continues to optimize those.
The programming team is growing by recruiting engineering students, just one or two years out of school, and training them on the principles of energy management and how to use the proper tools.
The sophistication level of the chiller energy management system is quite high as users acquire real-time input data directly from sensors outside, on the chillers, and in the buildings. The team conducts a series of variance calculations of the real-time input data with proportional-integral-derivative control loops, and then uses this data to determine and send new operating instructions to the chillers using small electrical signals. There are no moving parts, and team members are able to determine operating instructions using a series of genetic algorithms that combine heat transfer principles, thermodynamics, and advanced mathematical predictions, creating a “DNA footprint” of the entire system.
With the graphical interface, users can literally see the heat flow coming into and out of each building. Another benefit of this graphical real-time monitoring approach is they can cost-effectively handle the user’s energy management remotely.
The facility manager for a hospital or the plant manager for a manufacturing plant is responsible for the entire facility. These people are experts in different areas and remain concerned about efficient production and making sure nothing breaks down. Although they are under pressure to reduce costs through energy savings, they do not have the time or the experience to monitor energy usage and conserve energy. As it turns out, centralized chillers are the largest power consumer in a facility. Statistically, 45% to 60% of operational costs come from the centralized air conditioning, and 40% of that is from the chillers alone. The goal for the system is to handle the energy conservation for the chillers.
After the systems go in, engineers implement real-time monitoring from their offices to monitor and control systems anywhere in the world. The system modulates the chillers according to the change of internal and external load at 10-second intervals, automatically controlling the chiller loads. By benchmarking the average consumption of the chillers for each day, plant managers can see drastic changes in consumption when the system is in savings mode.
They record all necessary data and report it to plant managers to verify the amount of savings made in chiller consumption.
In the event of a mechanical problem, engineers can alert the facilities manager or the plant manager to switch over to their standby system while they make arrangements to make on-site repairs. The system monitors the number of hours each chiller has been running, and schedules preventive maintenance based on the manufacturers’ specifications for each chiller type and condition.
A hospital is one area the solution works, but the energy management system can adapt to other kinds of energy management, ranging from electrical power quality management, to decreasing the levels of diesel fuel consumption for boiler systems, to reducing electrical energy consumption through industrial air compressor optimization.
ABOUT THE AUTHORS
Joel Shapiro is group manager of Industrial Measurements and Control at National Instruments. His e-mail is email@example.com. Thirumalaichelvam “Thiru” Subramaniam is chief technical officer of Chiller Energy Management System (CEMS) Engineering in Malaysia.