The sceptic might be forgiven for believing that reliability in today’s consumerbased society means making sure that a device can be relied on to fail just after the guarantee has expired. This is a sorry state of affairs, but is in line with the unsustainable paradigm where economic growth is based on people buying consumer products they don’t need, which is to say that a healthy economy equates to unhealthy (neurotic) human behaviour...
Most readers of this article are fortunate in being able to work in infrastructure areas, for example, energy distribution, that genuinely aid the long-term development and well-being of society. This carries a significant challenge, because our view must be long-term and as comprehensive as possible. We need to predict the future, in terms of interest rates, load growth, fault frequencies, regulation models and future technology trends, while being aware of the limitations of our predictions. We also need to consider the past, by making the best use of the aging existing infrastructure and experiential knowledge we have inherited.
In the last century, developed societies have become increasingly dependent on infrastructure and in recent years reliability has become a major cost driver in network planning. It is well known, however, that increasing reliability increases investment and operation costs. By incorporating Customer Interruption Costs (CIC - in Finnish, KAH), something close to the optimum balance between investment, losses, maintenance and interruption can be found in network planning.
This leads to a dilemma. Generally speaking, it would seem good practice to head towards globally cost-optimum networks, but global cost-optimum, from a reliability point of view, may not be equally fair to all customers.
For example, providing a backup connection to a customer on the geographical periphery of a particular electricity distribution network might be disproportionately expensive. A globally optimum network solution might leave such a customer without a backup connection, meaning the customer, or cluster of customers, would have to wait for repair timebefore restoration of power after a fault in the line feeding them. If these customers are paying the same tariff as customers with backup connections, they could justifiably say that, over the long term, they are not receiving fair treatment.
Considering network planning a bit further, we have found that the CIC values assigned to the various customer groups have a large effect on the topology of a distribution network. If they are overvalued, there will tend to be overinvestment due to a greater number of feeders and backup connections, and more investment in switching. Interestingly, making a distribution network more reliable for customers means having more network, which means more faults - but shorter interruptions.
It is difficult, perhaps impossible, to put monetary values on everything, but quantifying the cost of interruption has brought reliability firmly into network planning. The same could be said for including costs that represent environmental impact in the cost functions in planning algorithms, and even more so, in the full life-cycle costing of all consumer products...
Robert John Millar is a D.Sc. (Tech.) and researcher at the Aalto University School of Science and Technology. His research interests include network planning and cable rating. How John got to Finland from his native country of New Zealand is a long story, involving music, love and the sense of space. He has been here, most of the time, since 1992. John likes the sauna and skiing aspect of life in Finland, and can still be heard in various bands playing the clarinet and other more exotic wind instruments.