If you have followed this series and the provided links, then you know that:

  • For avoiding storm caused tree-related outages, there are only three possibilities: remove overhanging branches; decrease the electric system’s tree exposure; and possibly, where the VM program has been grossly underfunded, increase the intensity of the hazard tree program
  • Increasing the frequency of pruning will do little to improve reliability
  • For a best-in-class VM program, the majority of tree-related outages are due to the failure of apparently healthy trees (danger trees) located outside the right-of-way
  • During major storms, the influence of hazard trees on total tree-related outages may be statistically insignificant
  • Hazard trees being more vulnerable to failure, the benefit of the hazard tree program is realized during normal operating conditions
  • The change in the VM workload can be described by a logistic function, which informs us that underfunding the VM program will result in exponentially expanding tree-related outages and ultimately, costs
  • The impact of underfunding may not become apparent as deteriorating reliability for the first three to five years, implying an appropriate window for evaluation is at least ten years
  • While we haven’t specifically discussed this, given workload expands by a logistic function, an asset management strategy where remedial action is triggered by the frequency of failures, is a strategy that imposes higher-than-necessary VM costs and tree-related outages on the customer
  • The VM workload and tree-related service interruptions can be held steady or in equilibrium by removing the annual workload volume increment
  • There is an optimal maintenance cycle based on the utility and community accepted clearance standards for each VM operational practice and average tree growth and mortality rates
  • The annual workload volume increment and, therefore, maintenance cycles can be substantially altered by drought, pest infestations or periods of above normal precipitation
  • As trees adjacent to power lines keep growing, the risk of tree-related outages is gradually increasing over time

In this context what are questions regulators should put to their utilities?

  • What percent of uncontrolled outages are due to trees?
  • What is the 10-year history of tree-related outages?
  • What percent of tree-related outages are due to branch ingrowth?
  • What percent of tree-related outages are due to branch failures?
  • What percent of tree-related outages are due to whole tree failures (uprooting, trunk break)?
  • What percent of tree-related outages are due to the failure of apparently healthy, structurally sound trees?
  • Where are the trees that cause outages located? Inside the right of way? Outside the right of way? How far from the nearest conductor?
  • What is the total danger tree exposure or population outside the right of way?
  • What is the proposed budget for VM? How was it developed? Historical costs? (May be OK but likely underestimates the amount required for hazard tree mitigation)
  • What is the annual workload volume increment?
  • Are there quantitative expressions of risk associated with funding VM at levels below the proposed budget that can be used to inform decision making?
  • How is the proposed VM budget related to reliability? Are expenditures focused to where there will be the greatest return in reliability for the dollar expensed? If not, what is the justification?
  • What percent of the budget is dedicated to reactive (hotspot) work rather than to planned work?
  • What are the maintenance cycles for the various VM practices and how were they derived? (Any utility that tells you that they do not have cycles, likely has a high incidence of tree-related outages; is using a reactive strategy that is costing customers both in money and reliability of service; provides little to no credence to their VM professionals)
  • What is the endemic tree mortality rate? Has a protocol been developed to measure changes in mortality due to drought or pest infestations?

Recommendations:

Do not allow VM funding simply to be hidden within O & M expenses. Require a separate reporting of VM expenditures. This greatly reduces the risk of senior utility management derailing the VM program by diverting funds to other areas and thereby provides the VM professionals an opportunity to perform.

If trees are the number-one or -two cause of unplanned outages, regulators would be well served by obtaining VM expertise for performance management audits, storm investigations and rate cases.

Secondly, have the top utility VM professional appear before hearings. If utility management resists such a request, it is unlikely that they will be able to cogently answer questions such as those above and it should become clear that the hearing outcome will not be favorable to them. On the other hand, be prepared for the fact that utility management may pressure the VM professional to attempt to justify their intentions or past actions, for example, such as having gutted the VM program budget. Again, the answers to the questions provided above should provide insight into whether the VM program is established on a solid foundation or is continually subjected to changes in course, pulled in a multitude of directions due to the influence of senior management, engineering, asset management departments, community groups or other state agencies. Even if the company provides unencumbered access to its VM professionals, you may not receive what you consider satisfactory, elucidating responses to your questions. I have often heard VM professionals bemoan the fact that the program is underfunded. Clearly, they feel that is not the time to make a case for a study that will establish program resource needs. When funding is considered adequate, then there appears to be no need to gather such data. However, the more penetrating the questions encountered in a regulatory proceeding the clearer it will become that there is an expectation for data that informs decision making and results in improving service, preferably while reducing costs.

Do not specify maintenance cycles. You will eliminate the option of doing only the work required and thereby, pass on to ratepayers expenses which do not have a tangible benefit. Do specify a minimum clearance between trees and conductors and set penalties for violations. You can find examples in California, Oregon and the province of Alberta. The regulations in Alberta are particularly interesting. The Alberta Electric and Communication Utility Code, which specifies minimum tree to conductor clearances in section 18(1), falls under Occupational Health and Safety Act or labor code. This creates a unique situation in that the responsibility for maintaining tree to conductor clearances applies not just to the utilities but also to landowners. VM professionals are greatly supported by minimum tree to conductor clearances as management does seek to avoid violation of laws and regulations.

Where excessive tree-related outages have been successfully addressed by actions made possible through increased funding, do not take this result as an indication that the work is done and expenditures can now be cut. It’s possible but without knowing what the annual workload volume increment is that needs to be addressed, cutting funding is almost certain to begin a new cycle of exponentially expanding workload and tree-related outages. For a utility that doesn’t know what the annual workload volume increment is, asking them to justify on going VM funding even though reliability has improved, is asking the utility to prove that the outages that did not occur, would have occurred but for the ongoing expenditures.

Do not ask for benchmarking. Unless you have extensive utility VM expertise you will not be able to distinguish whether you are comparing apples to apples or oranges. Consequently, it is highly unlikely that benchmarking that you ordered to inform your hearings will lead to decisions that improve reliability or sustainably decrease costs. I will explore this topic in detail in another series of articles.