CAN Newsletter September 2006

Chiller plant control and efficiency optimization

Energy costs represent a major expenditure in most companies. With the ever rising cost of energy and ever declining availability of fossil fuel, any reduction in energy usage will have a significant positive impact on profits and improve the competitiveness of business organizations. This concern is the main driver behind the move by many building operators and owners in Malaysia to gravitate towards sustainable solutions that offer long-term energy savings.

It has long been realized that in order to achieve significant reductions in energy usage, the emphasis should first be focused on equipment that consume large amounts of power. Studies conducted in Malaysia show that a large proportion of operating costs (between 40% to 60%) in commercial buildings is contributed by air conditioning equipment. Since it is very common for commercial and industrial buildings in Malaysia to have centralized air conditioning systems, the need to effectively control and optimize the efficiency of air conditioning chiller plants in buildings is of prime importance.

A system of this nature was recently implemented for a large telecommunications service provider in Malaysia by Controller Area Network Solutions and Energy IR. The end-user views this system as a solution for long-term reduction in energy costs. They decided to implement this system with the realistic goal of achieving a reduction in overall energy costs by at least 15%. The system was implemented in several buildings they owned and the results to date have been very encouraging with cost savings within expectations. With the implementation of this system, the end-user is able to remotely control and monitor the air-conditioning chiller plant through a process visualization system. The system can also perform automated control, with an emphasis on energy efficiency, of the chiller plant based on various parameters such as daily load profiles, outside air temperature, building air temperature, water pump speed, cooling water flowrate, etc. Remote alarm management is also incorporated such that the building maintenance team is able to respond immediately to system anomalies. The collected data also forms the basis for analyzing and implementing future enhancements to the chiller plant. Hence it is an important tool, which needs to be accurate and stable at all times.

The CAN protocol was adopted as the protocol of choice for this application. The main reason for choosing CAN was its suitability for use in a distributed network architecture. At the same time, its multi-master hierarchy, sophisticated error detection and retransmission mechanism also provided a high level of data integrity, which was crucial to the client. The application layer protocol conceived for this application was a proprietary protocol developed by Controller Area Network Solutions and was aptly named Coffee (Communication Fieldbus For Energy Efficiency). One of the key enablers of the system is the Control and Monitoring Module (Camm), which is physically dispersed throughout are the building and the chiller plant. Each Camm is connected to the Coffee network backbone. The Controller Configurator Software is used to configure and parameterize each Camm controller through the network. Each controller is given a unique identifying ID and communicates with an Application System Server that is also connected to the network. The server acts as a data collection and processing host. It contains a variety of software, including an energy consumption calculation software, SQL database, report calculation software, web server, SMS gateway software and a web-based process visualization software. The web-based process visualization software may also be accessed from a remote location via the Internet. It is important to note that although many common commercial off-the-shelf solutions were available to the system integrators, such solutions were not able to meet budget considerations and lacked features required for this specialized application. Since Return on Investment (RoI) is always measured with capital costs as a basis, it was essential to keep the capital costs as low as possible to maximize RoI for the end-user. At the same time, issues such as accuracy of measurements, temperature drift, the ability to perform PID calculations, I/O mix, physical size limitations, power supply consumption, reporting capabilities, energy efficiency calculation capabilities, etc. were important technical considerations which made common commercial off-the-shelf solutions infeasible for the budget available.

XML device descriptions for CANopen files

After years of standardization the new CANopen device description format according to ISO 15745 [1] awaits release. The advantages of XML in general — its extensibility, the support by a huge number of tools, powerful extensions such as XLST, XPath and many more — have already been discussed in many other articles. E.g. in [2] its advantages for industrial automation are explained in detail. Thus this article focuses on the user’s view of the CANopen device descriptions without extensive introduction into the format.

New features of the CANopen device descriptions

The device descriptions according to ISO 15745 consist of a part to describe the application (ProfileBody_Device) and of a part for the communication (ProfileBody_Communication) - in our case CANopen. Consequently the device descriptions in XML allow a description of the application and the device functionality that exceeds the plain list of objects in present EDS files. Furthermore these files support internationalization. Just one device description file is necessary even if the description is adopted to regional markets by adding descriptive texts in respective languages.
The description of the application in the device description file is an informational gain compared to the old EDS files that it is worth having a detailed look at it. The element ApplicationProcess within ProfileBody_Device is used to describe the application inside the device. It is subdivided in several lists to describe data types, functions and parameters and to describe their aggregation and instantiation. Templates may be used to combine parameters with the same parameters to reduce the file size. This functional description shall be shown using a simple inverter as example. This inverter has one input parameter, one output parameter and one configuration parameter to configure the function (to invert or not).
It is now possible to describe this function by its input, output and configuration variables. Devices that contain more than one such function can be described using function instances and references to parameters. Redundant descriptions of functions and parameters can be avoided by using templates. This mechanism also helps to ensure consistency of data. Additionally, the XML device description may contain descriptive texts for each function element. Further it is possible that the output parameters of one function can be the input of another function and configuration parameters might be used for various functions. So the new device description is a powerful way to describe the device behavior. The assignment of application parameters to CANopen objects is realized by references keeping a unique mapping of parameters to CANopen objects. Although it looks slightly complex at first sight this method is necessary for the separation of application and communication.

Single chip repeaters boost CAN applications

Controller Area Networks (CAN) have been around for some time originally having been developed in the early 1980s with automotive applications in mind. The CAN serial bus system protocol was standardized in 1993 (as ISO 11898) and is now available in hardware from numerous manufacturers around the globe for a range of applications that reach far beyond just automotive.
One of the original drawbacks of CAN was the limited physical distance over which it could operate effectively; CAN repeaters, as the newly released Amis-42700 , helped overcome this issue and also opened the door to many non-automotive applications such as elevators, industrial control systems and security systems where the physical distances involved were previously prohibitive to the use of the technology.

Historically CAN repeaters have been based around discrete component solutions. Now advances in semiconductor technology are making single chip CAN repeaters a real prospect. This brings a host of potential benefits that include performance improvements, component count reduction, power consumption savings, cost savings, and better resilience to harsh environments.

CAN background

CAN is a serial communications protocol which supports real-time control and multiplexing with a high safety level. It provides two communication services: the sending of a message - known as data frame transmission, and the requesting of a message – known as remote transmission request. Other services such as error signaling and automatic re-transmission of erroneous frames are performed automatically and are user-transparent.

Within a CAN system a multi-master hierarchy exists, this allows the building of intelligent and redundant systems meaning that if one node is defective the network continues to work unimpeded. A further feature is that information sent is transmitted to all devices on the bus, all receiving devices read the message and then decide if it is relevant to them. This along with error detecting mechanisms and faulty message re-transmission guarantees data integrity.
Repeaters are a vital part of a CAN system. The Amis-42700 semiconductor can be used to perform roles that include:
Enabling the network to cover greater physical distances

With increasing electronic content in passenger vehicles CAN network lines are getting longer. Industrial applications such as elevators can potentially have networks with lines of between 100 m and 200 m in length.
Maintaining impedance levels for a network at the interface with diagnostic devices and other equipment that is regularly connected and disconnected

For example connecting diagnostic equipment directly to the CAN bus, eliminating the controller in-between, improves communication speed, diagnosing, and operation. Another similar example is in a truck and trailer application adding the trailer doubles the length of the CAN line. A repeater used at the interface keeps the impedance the same whether the trailer is connected or not.
Based on discrete component solutions, CAN repeaters used in most current applications comprise a module containing CAN transmitters and receivers, a micro-controller and relevant supporting logic. The technology is well established but in the latest applications its larger form factor a lack of robustness are being exposed when compared to new approaches.