So is there a right maintenance cycle? Yes, there is. However, determining what that is requires a blend of quantitative data and qualitative decisions. In geographic areas vulnerable to wildfires, the pruning cycle should be such that tree-conductor contacts are avoided. Where wildfires are not a concern, public safety and reliability become the drivers of the pruning cycle length.

My personal view is that when in-growth outages account for 10% or more of all tree-related service interruptions, then you do not have a least cost sustainable VM program (see information about hotspotting, pg 32, "TransAlta Utilities Reporting System—A Management Tool"). As the majority of tree-related interruptions arise from tree in-falls, the hazard tree cycle length has a greater influence on reliability than the pruning cycle. As the entire electric system cannot be monitored in real time for hazard trees, there is always a residual hazard tree population. Utilities are faced with a qualitative decision regarding hazard trees, having to choose a maintenance interval that balances reliability and service costs. The shorter the maintenance cycle the greater the cost. However, the utility is left to decide whether the ratepayer is well served by the reliability gain available for a shortening of the maintenance cycle and the attendant increase in costs.

The percent of in-growth outages are one indicator of the appropriateness of the pruning cycle length. The other means of assessing the pruning cycle is via field inspection. This inspection requires an understanding of what situations represent a safety and/or a reliability risk and when this risk will be realized. Utility best-management practice is to prune for the maintenance cycle duration, not for a specific clearance between conductors and trees. That is, the clearance provided is species specific. Consequently, the inspector must be able to predict average regrowth rates for each species and situation. To assess the pruning cycle, the inspector must also be knowledgeable about how trees cause outages. For example, where it can be anticipated that trees will make contact with the neutral of a 12 kV line over the next growing season, this does not represent a risk to reliability and represents a very minor safety risk. If we assume a horizontal construction and the trees are approaching and will make contact with the primary over the next year, under dry conditions this is a fire hazard but it constitutes only a minor human safety and reliability risk. When in-growth outages account for 10% or more of tree-related outages, this will be evident in the field as many locations will be seen where branch growth is already capable of bridging phases or it will soon do so.

An evaluation of the appropriateness of the pruning cycle and indeed the maintenance cycle length for all VM operations is a proper and potentially rewarding line of regulatory inquiry. However, if it is to serve the ratepayer it must be grounded on a comprehensive knowledge of utility VM, not be based on wishful speculation. Shortening the maintenance cycle does not necessarily improve reliability but it does increase VM program costs.

The proper maintenance cycle performs work before it becomes necessary to use a more expensive method, such as allowing brush that could be treated with herbicides to grow too tall and require mowing or allowing mowable brush to become so large as to require removal by chainsaw (see Figure 1 pg 5). On the other hand, the proper maintenance is just-in-time from a reliability and safety perspective.

What is the potential impact of setting the right maintenance cycle and conversely, the implied impact of not getting it right? In the province of Alberta, I have reported to the Alberta Utilities Commission that the average cost per hectare for AltaLink’s VM program has been reduced 49.6% since 2005 as a consequence of increased funding that permitted adjusting the maintenance cycles. Obviously, the maintenance cycle is an important determinant of VM program costs. However, being able to deliver the optimal maintenance cycle is dependent on program funding.

In Part 4 we will discuss program funding, why it is possible to have a VM program that simultaneously delivers the lowest possible frequency of tree-related interruptions and the lowest costs whereas underfunding will lead to exponentially expanding tree-related service interruptions and costs.