Hydroelectric power stations produce the vast majority of Sweden’s production of electricity.* The system for production consists of reservoirs and dams in the upper parts of Sweden’s network of very large rivers. Reducing the upper water levels in the dams produces the static energy used to generate dynamic energy. Turbines convert this dynamic energy into electricity via electrical generators. After being transformed into an appropriate voltage, the electrical energy is distributed via a vast network of power grids throughout the country. In this way, the electricity is carried from the remotely situated dams to populated areas and industries. Until fairly recently, the cabling was mainly mounted on very high-level pillars. Nowadays, the grids nearby populated areas are channelled mainly underground. High volumes of electricity are distributed via underwater cables along the coastlines or across the seas.
Power generation—including the dams, dam gates, turbines, and electrical generators—is supervised and controlled from special control centres, including control rooms manned by people. In the past a large proportion of the control and logistical
manoeuvring was done manually. Successive technical devices have taken over most control functions. Mainly responsible for routine monitoring of the system, the operators have a key role in intervening in order to avoid malfunctions and accidents. Over the past decades there has been a separation of the control rooms for function of electricity production and the control of electricity distribution.
A central control room monitors and controls the overall production and distribution of electricity, including the import and export of electricity for the whole country. This centralised national control centre could handle (at least in theory) all of the monitoring and control functions for the whole country. In practice, it is thought wise to maintain a number of regional control centres to supplement the local centres and the control rooms for the production of electricity. In 2007, there were only three regional control centres. In the future these might be reduced to one or two.
In the past the people working in control rooms and at the local control centres placed a strong emphasis on the practical value of having personnel close to key parts of the distribution grid, including the power production control rooms. Local knowledge, in the form of tacit knowledge, is of great value to reduce disturbances when there are major fluctuations in electrical supply, such as during large storms. When parts of the distribution grid are disrupted and become incapable of electrical distribution, there is a risk that the problem will create a local backlog. This could result in an overloading of the remaining parts of the network, and might disrupt even larger areas of the network. It is not difficult to imagine the knock-on effect from a minor problem spreading into larger regions, causing large numbers of people and organisations to be without electricity.
Thus one important area of technical and human-technology risk management is to define the optimal combination of the automated system in relation to human operator support. In the past, a weakness of the system has been in fully understanding the actual capability of the human operators. In this context, a classic dilemma is local knowledge versus centralised decision making.
The enormously large dams represent another area of technical risk management. In fact, all of the different types of related technologies needed to generate and distribute electricity involve degrees of technical risk. The automation system as such also represents a degree of risk. The related computer system and its software is, to an increasing degree, an important risk factor. To cover this part of the system, an appropriate risk management capability is needed. This capability also includes available skills and knowledge among the control room operators.
Other sectors of the ‘control room industry’, such as the paper and pulp industry, are considering the use of operators specialising in software engineering on twenty – four-hour duty rosters.
In the past control rooms were more or less allocated near to most power generation sites. Nowadays, it is only necessary to have a few control rooms cover a large number of power generation units and related dams. One consequence of this development is the obvious reduction in the access to local knowledge among the operators. One control room might control an area of many hundreds of kilometres. An additional problem is that supplementary human visual contact is needed to ensure the more direct supervision of dam gates and other areas. In this new context, supervision needs to rely on supplementary use of locally placed video cameras connected to related displays in the control rooms. Today, the quality of video display techniques is of rather high quality, but there remains the issue that the video images are two-dimensional whereas the real world is at least three-dimensional. The fourth dimension (time) has also to be taken into account. For example, an operator might have difficulty perceiving slow changes in the environment such as a slowly increasing water level. Even in real life this can be difficult. The on-screen technical representation of the real-life environment might make the changing situation even more difficult to decode. This does not matter too much if the operators are well acquainted with the form and function of the real-world situation and have good local (that is, individual) knowledge. Thus equipped, the operators can to some extent mentally visualise the third dimension and add this picture to what they see on the screen. To some extent, they can also partly (but not fully) visualise the fourth dimension, time.
An additional problem is that there is also a need to add on different forms of security displays to control illegal entry via gates, doors, and other access points. A big difference today compared to previously is that there are increased numbers of video displays to handle the control and supervision of all stages in the power generation process. One reason for these changes is the increase in the numbers of threats by groups with a politicised agenda. However, when localised (individual) knowledge is missing, the task becomes even more difficult.
Some decades ago, research into control room work and automation involved intensive debate about the need for an overview of the function of the different subsystems in relation to the overall system. The old types of control panels frequently consisted of functional drawings of the related process, including analogue (frequently online) instruments for important variables and also some controls (knobs and dials for making adjustments and on-off switches). The positioning of the controls, visible to the operators, meant that the controls as such also became displays of the current control status. In the era of intensive computerisation of the control process, the old fixed panels were removed. The old inflexible hardware was replaced by keyboards and cathode ray tube (CRT) displays. Obviously the new solutions were much more flexible and thus it was much easier to update the changes in the process.
However, at the same time, the advantages of having a good overview of the system were lost. A major problem is the fact that the display modality of control positioning has disappeared. This means that the operator receives no immediate visual feedback from the positioning of the controls’ (knobs and levers) current status. A striking example is the use of a manual gear stick of a truck. By looking down at the position of the gear stick the driver receives immediate visual feedback. For a practiced driver it might be sufficient simply to feel the position of the gear stick. This kind of visual feedback is lacking for the operator who is only using a keyboard. This is the negative functional trade-off from the flexibility offered by the keyboard. In the future this deficiency has to be resolved. It might be difficult to return to the old types of controls in the forms of knobs, dials, levers, and so forth. But it might be useful to include into the display the control positionings or control status. In some special cases it might even be prudent to use ‘old-fashioned’ types of controls.
In many modern control rooms (not only in power control rooms), the design of the information presented to the operator seems to distance itself from the real-world process that it should represent. It is important that the information presented gives the operator the correct and optimal feeling for the process and its current status. This should be the main role of an overview process display. This display should continuously support and update the operator’s mental and tacit model of the ongoing real-life process.
It is a risk to put too high a degree of emphasis on local desktop displays. The desktop should give detailed information. Frequently, an operator works with several desktop displays in parallel. It is important that the different displays represent different modalities. The overview display should show a representation of the status of the main process.
Overall, existing displays seem to be orientated more towards rather static information in the form of different types of important indicators—for example, numerical. Ideally, such information should be process and dynamic orientated in a graphic image form. New display techniques for macro presentations of overall functions of processes could probably improve the operators’ understanding and ability to intervene rapidly and appropriately. Investments in these types of new displays are most likely very cost effective.
Earlier in this handbook, Figure 6.14 showed an example of a layout for a control room. A workplace is placed centrally with access to a display unit, keyboard, and special controls. There is also space on the central workspace for writing and for various reference books. The process, which is assumed to be large, is displayed on the overview panel, mainly in the form of static information. Inserted into the display is certain dynamic information, such as the settings of the more important controls, critical readings, or deviations from the set values. The overview board is designed so that all parts of it are at more or less the same distance from the operator. There is also a more traditional office workplace in the control room, where various special tasks can be carried out during daytime.
In a control room where there is an overview panel and several people are working, it is important that this is positioned so that everyone can see it.
A later section (Chapter 11) of this handbook describes the design of learning and creativity at work. In most control rooms for power generation the operators are underloaded. In such a situation, the work tasks do not prepare the operators for the more demanding tasks that occur, albeit infrequently. There might be hours or days between major highly demanding tasks. The main reason for having operators in a control room is to handle these high-peak critical tasks. It is important for the operators to prepare themselves for these essential situations. One way is for the operators to use the existing computer systems to simulate possible breakdowns or other types of critical incidents. This will allow the operators to improve their ability in handling disturbances. They will also obtain improved knowledge and insight into the function of the control system and the processes controlled by the automation. As a part of this simulation training, the operators themselves can contribute to the development of the system and its automation. If the operators become proactive in the development of their own workplace and its related tasks, they will also become much more motivated and creative.