Mike musings: short term plans and long term consequences

 

The great event of the 2016, namely the publication of the IRP 2016, has finally happened, and the plan has received both praise and criticism from many quarters. The media briefing went into some detail of how the plan was constructed and all the factors taken into account. Without discussing the plan itself, I would like to consider the complexity of compiling such a plan, which runs up until 2050, some 33 years in the future. One could ask why it is necessary to look so far into the future, as uncertainty increases with time. In fact one commentator made the statement that the only certainty in long term planning is the laws of science. The construction of such a plan must be a daunting task, as both over- and under-provision are equally critical.

Long term planning for demand and capacity has come under question in the last year, and the current popular opinion is that as we cannot predict demand in the long term and should change our plans to follow short term trends. Thinking seems to favour short term plans that could follow demand more closely, and which would be based on smaller distributed generation plant close to load centres rather than large centralised units. This is probably influenced by the experience of Medupi and other large systems.

Load-centric distributed generation and short term decentralised provision sounds like a wonderful idea and has been proposed by many sources, but has not been studied in detail. Bar a few studies aimed at highlighting the benefits of distributed generation, there does not seem to have been any detailed studies done on the subject. Even the definition of distributed generation is rather vague, and the long term effects of short term planning and short term provision have not been investigated.

One of the items that has not been considered is the impact of build times of smaller plant. The assumption is made that “standardised” designs could be used to reduce build time, but there are other factors affecting the time to complete. For instance how long does it take to construct the equivalent energy generating capacity of a 3000 MW conventional power station. If we assume a unit size of 500 MW say, it would require the construction of six units at separate sites. If we assume 100 MW this would require 30 sites. If we assume renewable energy as the technology it would require approximately 100 sites of 100 MW baseplate capacity. Who would decide where and when to build these smaller plants?

One of the issues of course is how to control and manage such a short term programme. No matter how distributed, load centric and right sized power stations are, they will accumulate over time and could result in an unmanageable mess, which defeats the object completely. The first issue of course is who would do the planning? Would it be centralised or localised? Under the present regulations, only the DoE can decide on additions to the utility generation fleet. Would centralised planning work for distributed generation?

One of the other concerns is that a large amount of distributed generation plant concentrated in one area could begin to behave like a single centralised plant. Are we not heading in that direction anyway with the REPPPP? All the planned RE systems are connected to the grid, and are located at sites based on generation capability and not proximity to load centres. At the moment, the REIPPPP is not area specific and choices are based on the bid price. So we could end up with a large concentration of plant in one area.

Concentration of RE in specific areas such as the REDZ, amounts to centralised generation does it not? I we also consider that the plant is self-dispatched, with all the solar or wind farms producing at the same time, we have an effective centralised generation plant. Individual plants are not despatched or controlled from a central point.  It is agreed that geographic diversity smoothes out some the short term variations in output, but the gross effect is still there. This is equivalent to smoothing out the ripples on the surface of a tsunami, but the wave is still there.

A quick glance through data provided by the CSIR on wind patterns shows that in spite of claims that the wind is always blowing somewhere, there are many occasion when the wind is low everywhere in the country and likewise, when it is high everywhere. Gross effects apply over the same geographic area and plant located close together will be subject to the same gross effect. Are we are heading for the equivalent of unplanned centralised generation?

While we are on the subject, why is the REIPPPP technology specific? Surely a reasonable program would be to specify demand and time, and let the industry sort out how to achieve it. Should the future planning not follow this path?

How much of our thinking is based on what is available at present, and immediate benefits without considering long term performance, and projected into the not very distant future.

Maybe our thinking is influenced by the short lifetime of other things in our lives. Very few people drive a car for more than ten years, a computer more than ten years old is obsolete, box type TV has been replaced by flat screen. Is it the fact that we know that technology will change in the long term and therefore planning for more than a short period is meaningless? All these short term items are easily replaced by newer ones, but the same cannot be said for infrastructure. Power stations do not disappear, they accumulate.

Short term choices can have long term consequences, especially when the subject of the choice reaches its life end and another choice has to be made. If no replacement is available or the market is limited to one choice, one faces the option of having to start all over again.

If we are to adopt distributed generation, the whole process needs to be looked at carefully.

Most definitions of distributed generation take the closeness to load centres into account rather than the price. The current REIPPPP only takes the cost of connection to the grid into account for pricing. Maybe a future criteria of the cost of delivering energy to load centres would give a better picture and result in a more distributed nature to the future grid.

Send your comments to energize@ee.co.za

 

The post Mike musings: short term plans and long term consequences appeared first on EE Publishers.

Source: EE plublishers

More news