Home > Featured Articles > Initial results of Unitil’s Storm Resiliency Program

Initial results of Unitil’s Storm Resiliency Program

By Sara Sankowich


Utility companies employ vegetation management programs to minimize outages during normal days and minor weather events. A typical program involves assessing circuits and removing diseased, weak or dead limbs, and trees that are encroaching on utility lines or poles on a scheduled cycle (for Unitil, this is every five years). While this program has proven effective in minimizing outages, a recent increase in storm frequency and intensity in the Northeast has tested typical vegetation management programs to their fullest.

As a system arborist for Unitil — an electric and natural gas utility serving Maine, New Hampshire and Massachusetts — my responsibility is to ensure vegetation adjacent to the overhead conductors is managed in a manner that maximizes Unitil’s resiliency and safety, and minimizes customer outages. During the past several years, these more frequent and powerful weather events challenged the systems we had in place. In response to public demand, we put a more proactive program together to reduce the volume of outages from tree and limb failures during major storms. Early results indicate that the program is off to a great start.


Unitil’s Storm Resiliency Program is born

In 2012, Unitil implemented the Storm Resiliency Program (SRP) as a companion to the existing, more traditional core vegetation management program. The crux of the SRP is as follows:

  • The program is performed only on select circuits, and then only on the critical portions of those circuits — typically the three-phase from a substation out to a desired protected device, with the goal of protecting as many customers served on the line as possible.
  • Unitil arborists and trained inspectors do an intensive risk tree assessment of all trees tall enough to fall and impact the electric lines, with the goal of removing trees that pose a risk of failure due to a defect, disease or species-specific failure trait. For trees that do not pose a failure threat, the goal is to perform ground-to-sky clearance or to remove all limbs overhanging the conductors.

Local first responders are consulted to target areas that will improve lifeline, life safety, and communication resources. Those crucial services — police, fire, schools, shelters, hospitals, gas stations and food — along a critical path will get first priority in SRP scheduling.

The SRP’s kick off began with an intensive assessment in the Atkinson/Plaistow region of New Hampshire in 2012. This started with a manageable number of miles to review and mitigate to gauge customer acceptance, effectiveness, and cost.


Crane Crossing copy

Before SRP (above) and after (below). All photos and graphics provided by Unitil

Testing the system

A new program isn’t worth its salt if it can’t handle a test or two, and Unitil’s SRP is no exception. Mother Nature dealt the first test early on when Hurricane Sandy blew through in October 2012; one circuit of storm resiliency work was nearly complete, and work had just begun on two other circuits. Although the region did not receive a direct hit, the strength of the storm still caused approximately 440,000 customers to lose power in New Hampshire and Massachusetts. Interestingly, the circuit in which SRP work was almost complete did not suffer a circuit lockout (meaning 100 percent of the customers served on that circuit are without power) during the storm. Although not on the main line, one tree failure occurred on this circuit’s SRP area, and, at that time, we were already in negotiations with a hesitant customer about removing the tree when the storm came through. On the other circuit, which had already been inspected, two failures occurred on our main line where we had marked trees for removal, but had not yet removed them; as a result, the circuit suffered a lockout.

Crane Crossing 3yrs copyStatistics at that time proved that the program was making an impact. We used the almost complete circuit to estimate the number of events we may have avoided from doing the SRP work. Looking at the data, only one event occurred (due to customer refusal) in a 5.5-mile area where we had performed resiliency work, while the remaining 20.5 miles of line on that circuit experienced 18 events. This means that the portion of the circuit that did not have storm resiliency work had an event every .7 miles. If we hadn’t done the SRP work and had an event every .7 miles on that SRP area, an additional 3.85 events should have occurred — and had not.

Another early indicator that the program was shaping up to be a success was customer reactions.

In the years that followed, we took those early results to the state of New Hampshire and asked to move the program from pilot status to permanent, and then later proposed it to Massachusetts as part of our rate case. New Hampshire approved a second-year pilot, and then permanent program status, and in early 2014 Massachusetts also approved the SRP.

We kept plugging along, and soon thereafter, Mother Nature gave us a second test. On Thanksgiving morning of 2014, a storm blanketed New England with heavy, wet snow. Out of 30,000 Unitil customers in the Concord, N.H., area, approximately 23,000 were without power on one of the busiest entertaining holidays of the year. Three circuits in the Concord area underwent storm resiliency work in 2013, and were prime candidates to test the program. Two of those three circuits suffered no outages on their backbones. A third area suffered an outage caused, again, by a tree on private property we could not work on or remove due to customer refusal. While we weren’t avoiding all outages on those circuits, we were reducing the number of customers interrupted per overhead mile and the number of incidents in areas targeted by SRP work.


Customer education

Education has been an invaluable component in gaining acceptance to the Storm Resiliency Program. Unitil puts an enormous amount of effort into educating its customers on the impact of their diseased and dead trees on utility lines; however, we still occasionally receive push back. For example, several citizens in a small, picturesque town in New Hampshire questioned why we tagged for removal several trees they felt added to the character of their quaint downtown area. Because of our identification and recordkeeping process, and my involvement in the implementation, I was able to meet with those concerned and discuss the health issues tree by tree. In another town, a neighbor, not a landowner, objected to the removal of a tree because he would lose an important element of his view. Not only were we able to gain the approval of the neighbor, we were able to explain current limitation of the town’s tree care budget and help engage him to take proactive tree care steps in his town.

Unitil’s goal is to not decimate an area; instead, we take care to minimize our impact. We are able to leave surrounding vegetation undisturbed in some areas by using a crane to remove trees where possible. In addition, we work with landowners to improve the aesthetics after removal. We offer to leave the wood for the homeowner’s use, and encourage them to participate in our tree replacement program through the Arbor Day Foundation. The program allows customers to choose trees that provide energy savings around their homes; trees removed near wires can even be replaced in the same area with the choice of low-growing species.

At the end of 2014, our education efforts were paying off. We had only 11 refusals on the 2,000-plus trees tagged for removal in 2014, and although we weren’t able to remove the entire trees in most of those instances, we were able to perform some work on them to lessen their threat.



Training has proven to be the key ingredient in ensuring our vendors — two tree-removal companies and a contract forester used in New Hampshire — are performing up to our expectations. Forestry field supervisors typically spend at least two full days up front with the contract forester assessing and marking trees together and learning the program goals. We then review what has been marked in the first week of work planning and coach, adjusting the level of tree removal intensity as needed. We have a tree risk assessment training guide/field manual that contains a matrix and other tools to help us identify which trees to remove based on type and severity of defect(s), addition of modifiers, and likelihood of failure. Although modeled after the American National Standards Institute (ANSI) hazard tree assessment standards and fitting within the International Society of Arboriculture’s Tree Risk Assessment Qualification (TRAQ) program, a fair amount of subjectivity still exists within the program. The tree risk assessment protocol we developed helps to remove as much subjectivity as possible.


Results by the numbers

The goals in the first year were to ensure we could handle the additional workload, monitor the implementation process, and review customer perception. After the first year, and as the program went on, we continually assessed the program and the risk reduction needs. We left no prisoners in the first year — we covered 14.7 miles and removed 115 trees per mile. If a tree was a risk or if it had a defect, we took it down, all while continually assessing the trees that did fail on our system to learn more about risk and failure. During the second year, we more than doubled the miles, covering 32.3 miles and removing an average of 70 trees per mile. By the end of 2014, we completed storm resiliency work on 5.6 percent of our system, covering 89.6 miles and performing 6,502 removals.

The result of that work? In 2010, two years before the SRP began, we had 613 tree-related incidents, leading to 101,760 customers interrupted. In 2014, after multiple years of SRP, we had 383 tree-related incidents with 52,501 customers interrupted. The numbers indicate that Unitil’s storm resiliency and core vegetation management programs are not only protecting the system during major weather events, but also beginning to maximize resiliency and minimize outages throughout the rest of the year.


Comparison of Costs to Avoided Costs
Annual Component Cost Avoided Cost Cost to Customers (without the additional work)
Storm Resiliency Program $1,423,000
Major Storm Events* – $     76,972
Normal Operation Events – $   35,737
Public Direct Costs of Interruption Events $67,000,000
Totals $1,423,000 – $ 112,709 $67,000,000
* Assumes 1 major event annually


The future is bright

Our success thus far is encouraging for the future of the SRP. We plan to continue the program, measure results, and fine-tune the program along the way. We would like to narrow the results of the program to look more closely at the sections of SRP circuit worked, and also collect more tree-failure data in storm events so we can make a more informed appraisal of the SRP results. We are also looking at using our system data to identify the types of trees that are failing, and look for trends. At the 10-year mark, we will reassess, and if, in fact, Unitil’s Storm Resiliency Program continues to measure up to our ultimate goal of preventing customer outages in the wake of more powerful and frequent storms, and the public still supports it, we will continue.


Sara Sankowich is system arborist at Unitil, a public utility holding company, headquartered in Hampton, New Hampshire, that provides electric and natural gas distribution services in New England.

About jkmitta