T&D World Magazine

Vegetation Management Metrics

In February, I was at the International Society of Arboriculture Southern Chapter meeting in Savannah, Georgia. In a Utility Arborist Association session, we received presentations from three very experienced utility engineers. One of them, Greg Welsh, raised the topic of VM metrics.

This is a topic that needs far more consideration in utility vegetation management. In discussing it, I’m going to fire a shot across the bow of senior utility management, regulators, engineers and vegetation managers.

If you want to drive your vegetation managers crazy, have them respond to a benchmarking survey that contains only a half dozen questions--but does include the questions of how many years is one maintenance cycle and how many dollars are spent per mile. That should have the vegetation manager go apoplectic.

If that was fun, want see their hair go on fire? Have your VP walk in holding the results of that benchmarking survey and ask why your maintenance cycle is five years when a another utility that completed the survey has a 12-year cycle. Or why your costs per mile are higher than average and substantially higher than the utility with the 12-year cycle.

Are you a utility executive who has asked your vegetation managers such questions? The look of shock and disbelief on your vegetation manager’s face was justified. Dollars per mile is a horrific metric because the utility that most completely ignores maintenance is the winner. The customers of that utility may have many names for it, but “winner” is probably not one.

What about that darling of utility regulators, the maintenance cycle, or in their terms, the trimming cycle? The first problem with asking about the maintenance cycle is that there is no industry-accepted definition for the term. Further, other than my own writings in a performance audit of a Pennsylvania utility, I’ve never seen anything in print that attempts to provide a definition.

How can we use the maintenance cycle as basis of comparison if utilities have different interpretations of the term? Among regulators, there seems to be a broadly held belief that shortening the maintenance cycle will improve electric system performance during storms and consequently, some state regulatory agencies have specified a fixed maintenance cycle. How do the regulators define a cycle? Does it apply to how often the lines are patrolled? Pruned? Sprayed? Mowed? Examined for hazard trees? These distinctions aren’t made. Typically there is reference to the “tree trimming cycle.” What work is included in that? (See Vegetation Management Terms) In failing to make appropriate distinctions, regulators have utilities focus on work that will contribute little to improving reliability while potentially drawing resources away from precisely the types of work that could reduce storm damage.

The length of the maintenance cycle is only meaningful in the context of what result it produces. Without knowing the result, comparisons between various utilities’ maintenance cycles are more than useless. They actually do harm by misinforming and thereby, misguiding VM programs.

In Savannah, I raised the fact that dollars per mile was such a poor metric that I recommend its removal from the VM lexicon. There was complete agreement within the room. Why then do engineers doing benchmarking studies include this question? And why do vegetation managers measure their work in dollars per mile? I can appreciate that dollars per mile is useful in a historical context. However, as a vegetation manager, you need to appreciate that every time you mention this measure you give it credence and thereby, foster the folly of using it as a comparative metric. If you use it, how can you expect your VP to recognize it as an invalid metric outside your utility?

Is there a way to make dollars per mile a basis for comparison? Yes and no. If you sought first to group utilities by similarities in the tree species composition of the utility forest, the total tree exposure (both within and outside the right of way) per mile of line, the number of customers per mile and the percent of customer minutes attributable to tree-caused outages then dollars per mile could be used to compare utilities. In attempting to do so, you’ll find it practically impossible to form reasonable size groupings of utilities. In other words, it’s theoretically possible but practically impossible.

Regulators and utility management want to be able to measure the effectiveness of utility operations. In the field of VM, the metric I recommend is man-hours per work unit. The work unit may be a tree pruned, a 12-24” tree removed, m2 of brush sprayed, etc. as presented in Crew Evaluation by Robert Nosse. This approach removes the influence of non-productive time such as travel to the site and time spent dumping chips. It removes the influence of regional differences in wages. The work units must not only be similarly defined between utilities but must also be audited (See TransAlta Utilities Reporting System—A Management Tool Figure 6) to ensure proper and accurate reporting of work units completed.

Using the above mentioned approach, the effectiveness of each utility’s VM work could be calculated. However, this would necessitate a willingness to collect and audit the required data, industry agreement on work units, productive and non-productive items and work unit definitions. It would also require a body to collect data across utilities and to calculate the standard man-hours per unit so as to be able to make inter-utility comparisons of effectiveness.

We have no simple metrics for comparisons between utility VM programs. Using benchmarking metrics without an understanding of each metric and its proper context obfuscates rather than informs decision making. The perverse consequence of using improperly understood metrics is that regulators contribute to worsening reliability or excessive VM costs.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.