DCS versus SCADA: Everything in its right place

The debate on the advantages of SCADA (Supervisory Control and Data Acquisition) versus the advantages of DCS (Distributed Control Systems) has been running in Control Engineering now for several months. Here, Phil Gillard, General Manager of SolutionsPT, the UK and Ireland partner of Wonderware, calms the waters by arguing that each technology has its place.

Workflow software is a key part of Wondrware's system platform software

The debate over the comparative advantages of SCADA and DCS began when a Distributed Control Systems provider argued in these pages that the technology is now both cheaper and simpler to manage than SCADA. However, this misses a crucial point.

While DCS has evolved over the years and become cheaper and more intuitive to use, often looking and feeling much like a PLC interface, SCADA has also evolved. It has become more object-orientated, which means that it offers many of the advantages of DCS whilst retaining much of its own, traditional, flexibility. It also means that the lifetime cost of running a SCADA system has decreased in much the same way that the lifetime cost of running DCS has decreased. However, SCADA started life as a lower cap-ex solution, so is probably still more cost effective in the applications where it is more suitable.

The technical argument presented by the pro-DCS brigade is simple; with a distributed control system, you have only one database to maintain where as with SCADA you are likely to have more than one. If this is the only factor one takes into account, then the life cycle cost of a DCS is lower when one considers the total project lifecycle from system design through commissioning, training documentation and system modification.

The history of SCADA
However, one should also bear in mind that there are nearly forty years of misconceptions to battle in this space. The term Supervisory Control and Data Acquisition first appeared in print following the inaugural PICA (Power Industry Computer Applications) Conference in 1973. And since then it has become more and more complex and more misunderstood every year!

In order to establish the best use of each control medium going forward, I think it is worth briefly analysing the history of the two technologies. Distributed systems evolved from single loop controllers housed in a panel, which provided control and visualisation in the form of faceplates. In contrast, PLC and SCADA technology developed from banks of relays combined with visualisation.

Later PLC technologies replaced the relays with micro-processor based control. In contrast, DCS replaced both the control and the visualisation elements of the technology with micro-processors. The reason for this was simple, continuous control requires less processing power because the scanning speed is lower. As a result, there is the virtual memory and processing capacity available to fit visualisation into the same system.

This in turn led to the introduction of industrial PCs into the SCADA environment. However, because in the dark ages of SCADA technology PCs weren’t sufficiently robust for use in industry, the need for multiple databases arose. Modern industrial PCs have shed the stigma of instability and now offer a very reliable and stable solution for industry. The logical conclusion of this process was for DCS to develop as closed systems, while SCADA remained open and flexible.  

Single vendor benefits versus flexibility
In large scale continuous process plants, the control system is likely to be purchased as part of a turnkey package. As a result, a decision can be made as to which single vendor is to be chosen. This allows the end user and its engineering support organisation to take full advantage of the single database benefits of DCS. In doing so, they rest safe in the knowledge that the control hardware, visualisation and management tools will all come from the same development organisation; the DCS vendor. Because DCS systems were originally developed with large continuous customers in mind, it is highly likely that the functionality packaged within the standard DCS will be suitable for the application with little need for customisation.

However, the discrete industries, such as automotive, electronics, and packaging, and the hybrid industries, such as food and beverage, pharmaceuticals and mining, are an entirely separate issue.

These are the sectors where SCADA has its roots. In these worlds the systems tend to be more fragmented, often by the very nature of the process. In the case of the hybrid industries there is a need for a combination of continuous control, batch control and packaging. Equally, in the discrete (and packaging) industries there is often a fragmented supply chain where individual pieces of equipment are provided by different OEMs, who typically have their own standards for control hardware and local visualisation. The practical outcome in these industries is that there will be, by definition, a disparate number of systems and thus databases. As a result, the single database advantage of DCS ceases to be a real benefit and the flexibility of SCADA comes to the fore.

The question is how do we accept the realities of these two worlds and take advantage of the virtues of a single database? The object orientated functionality of modern SCADA is crucial here. Objects can be defined during the design process and managed through the system life cycle of commissioning, training, documentation and system modification. Furthermore, the scope of an object can range from a single I/O function to a combination of visualisation, historian and alarms, allowing the system to achieve objectives beyond those of a pre-packaged faceplate.

The key advantage here is that these standard objects can be applied independently of the existing control assets. For example, they can be applied to multiple types of control hardware within the same system, reinforcing SCADA’s suitability for discrete or hybrid applications. The technology has even evolved to allow automatic binding, through the project lifecycle, to the tags within the PLC.

SCADA in practice
An illustration of the benefits of object orientated SCADA is provided by Wonderware’s System Platform product, which is in used as a standard platform at multiple European airports. At the core of System Platform is something called the ‘plant model’; a logical representation of the physical processes being controlled and supervised. Object technology makes configuration, logging, delivery and maintenance of real-time and historical information in the airports relatively simple.

In many of these applications, they have the electrical and bridge systems, going to the aeroplane, as well as lighting, elevation and HVAC (Heating, Ventilating and Air Conditioning) in the building itself. Clearly, each of these disparate technologies has a separate control system, which they are able to integrate via SCADA and System Platform. This is something the airports are able to achieve using the flexibility of SCADA, but which would have been very difficult using DCS.

Another example that illustrates the flexibility of SCADA can be found at a renewable energy company with multiple sites across the UK. Again, the object orientated nature of SCADA delivers both operational and managerial benefits by allowing each site to be monitored and managed centrally, despite the host of control technologies and systems on display in each location.

These two case studies illustrate the fact that all engineering happens in real world situations, in which we have to cope with the realities of the multiple control systems imposed upon us by other elements of the end user’s supply chain. The reality of life is that in discrete and hybrid industries you don’t have the benefit of one vendor, one database solutions. 

DCS systems are still well suited to their traditional world of continuous process, whilst the world of SCADA has evolved to allow object orientated capability to reduce lifecycle costs in hybrid and discrete applications. Each technology has its own benefits and the best result for each application can be achieved by applying those benefits appropriately – it isn’t a battle. For me, the DCS versus SCADA argument ultimately comes down to nothing more than ensuring that each technology is in its right place.