Methods to Improve Herbicide Application AuditingIndustry
By Steve Mathews, AEP/PSO and Colby Norgaard, CN Utility Consulting
Herbicide applications have become an integral part of managing vegetation for many utilities. There are countless products, application methods and contractors available to help meet integrated vegetation management (IVM) goals. But after a utility chooses from the many product, application and vendor possibilities and begins to implement an IVM program, they often find that it is difficult to know if the program is meeting expectations. How much brush is there to control? Is the program effectively controlling the brush? Is it doing so within budget?
American Electric Power (AEP) Public Service Company of Oklahoma (PSO) determined that the best way to answer these questions was to audit the work being done and partnered with CN Utility Consulting (CNUC) to put together a program to measure results.
Utilities that have tried herbicide auditing programs know that it can be difficult to quantify the brush they want to control. Counting the number of stems is very time consuming and inaccurate because individual plants are extremely variable in canopy size, shape and height. Through trial and error, AEP/PSO and CNUC determined that measuring brush by square footage was the best way to quantify the undesirable vegetation around power lines.
Measuring rights of way covered by brush from drip line to drip line provided a well-defined unit to control. The square footage could then easily be classified as controlled or uncontrolled, and measurements could be quickly estimated. From there the total square footage of brush on the system could be determined, as well as the extent at which it was being controlled.
Random Sampling Using Spans
Our next step was to determine how thorough audits should be in order to ensure a successful program. One hundred percent audits proved to be accurate, but very time consuming and cost prohibitive, so we began exploring the application of random samples. This method allowed us to utilize the measuring techniques we had developed while being more efficient.
However, the process of selecting random locations along the system also proved to be time consuming. Spans between poles were determined early on to be more consistent than customer addresses in size and dispersal throughout the system. Locating, numbering and randomizing all of the poles on each circuit, however, took several hours, and finding each randomly selected pole’s location in the field added several more hours to the process.
The Plot Sampling Method
AEP/PSO and CNUC have more recently begun to explore the plot sampling method, which has been used in forestry production for decades. Plots have predetermined uniform sizes and are distributed evenly throughout the population area. As long as each plot is placed at the same interval, the results will be similar to random audits. Each plot can be measured and statistically extrapolated based on the data set. This method provides similar accuracy to a random sample, and it takes less time to select and find the random locations.
Auditing has become an important part of the IVM program at AEP/PSO. Random sampling with spans and plots both have improved the efficiency and cost effectiveness of our program. CNUC has also gathered valuable data that will help AEP/PSO improve its program in the future.