VIEWS: 0 PAGES: 14 POSTED ON: 10/5/2012
A QUALITY IMPROVEMENT “STORY” ABOUT PACIFIC ELECTRIC & GAS (PE&G) To illustrate a systematic approach to Quality Improvement, we will describe the activities of a quality improvement team within PE&G’s Haasville District. Our goal is to illustrate the following: The need to have a quantifiable measure of quality. The need for the quantifiable measure of quality to be customer oriented. How to use a Pareto Chart to select which of a problem’s “root causes” to improve. How to use a Fishbone Diagram to summarize a problem’s root causes. The importance of Knowledge Transfer and Learning. 1 STEP 1: Identify a Quantifiable Measure of Quality that is Customer Oriented The Haasville team first considers a goal of Provide Reliable Electric Service. Why is this goal inappropriate? “If you don’t measure it, you cannot improve it.” “In God we trust. All others bring data.” Next, the Haasville team considers measuring its service quality as the Number of Power Outages per Year. Although this measure is quantifiable, why is it NOT customer oriented? 2 Finally, the Haasville team decides that its measure of service quality will be the SERVICE UNAVAILABILITY INDEX. EXHIBIT 1: Summary of Computation SUI EXHIBIT 1: SERVICE UNAVAILABILITY INDEX Data for Most Recent Calendar Year Outage's Number of Outage's Affected Duration Outage Customers in Minutes Product Outage's Number (N) (D) (NxD) Cause 1 500 60 30,000 Birds/Animals 2 900 5 4,500 Lightening 3 300 15 4,500 Wind --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- 150 700 30 21,000 Lightening SUI = Sum of Products = 1,800,000 SUI is both quantifiable and customer oriented. 3 Having identified SUI as its measure of service quality, the Haasville team sets a goal of reducing SUI. The team examines each POWER OUTGAGE REPORT (POR). For each power outage, the POR includes: Time power went out and time it was restored, Number of customers affected, Cause of outage (if known) from a list of 35 possible causes. Using the PORs, the team identifies the root causes of power outages and summarizes them with the table below: EXHIBIT 2: Summary of Root Causes Number of Root Cause Occurrences Lightening 32 Birds/Animals 27 Wind 24 Construction 22 Vehicles 17 Unknown and Other 15 Trees 9 Equipment Failure 4 Column Sum = 150 4 What should the Haasville team do next? The problem with Exhibit 2 is that it does not display the impact each root cause has on SUI. Exhibit 3 below displays for each root cause both the number of occurrences and the contribution to SUI. EXHIBIT 3: Extended Summary of Root Causes Number of Contribution Root Cause Occurrences to SUI Lightening 32 315,000 Birds/Animals 27 126,000 Wind 24 225,000 Construction 22 180,000 Vehicles 17 540,000 Unknown and Other 15 45,000 Trees 9 360,000 Equipment Failure 4 9,000 Column Sum = 150 1,800,000 Although “Lightening” is #1 in terms of the number of occurrences, “Vehicles” is #1 in terms of its contribution to SUI. The Haasville team constructs the Pareto Chart in Exhibit 4 on the next page. The Pareto Chart displays the root causes of outages in decreasing order of their contribution to SUI. NOTE: To construct within Excel a Pareto Chart similar to Exhibit 4, start the Chart Wizard, click in Step 1 on Custom Types, and select Line- Column on 2 Axes. 5 EXHIBIT 4: Pareto Chart for Root Causes of All Outages 600,000 100% 540,000 550,000 90% 500,000 80% 450,000 Contribution to SUI 70% 400,000 360,000 350,000 60% 315,000 Contribution to SUI 300,000 50% Percent of Total 250,000 225,000 40% 200,000 180,000 30% 150,000 126,000 20% 100,000 45,000 10% 50,000 9,000 0 0% s s re s d n g er al le ee in in tio lu th m c W en Tr ai c O hi ni ru tF t Ve & gh A st s/ en n Li on w ird m no C ip B nk u Eq U Root Cause What should the Haasville team do next? 6 STEP 2: Analyze root causes of vehicle- related outages and determine which one to attack. Using the PORs that were vehicle-related, the Haasville team constructs the FISHBONE DIAGRAM in Exhibit 5 on the next page. The Fishbone Diagram organizes the root causes of a vehicle-related outage. NOTE: To construct within Excel a Fishbone Diagram similar to Exhibit 5, first display the Drawing toolbar by using the menu selection View, Toolbars, Drawing. Then, to draw the arrow-tipped lines, click on the toolbars arrow- tipped line icon. 7 EXHIBIT 5 FISHBONE DIAGRAM (CAUSE & EFFECT DIAGRAM) FOR OUTAGES CAUSED BY VEHICLES MAN High Speed Failure to Yield Incorrect Sag Driver Fell Asleep Construction Equipment l ho s Forced Off Road d ug co re Dr Al Ti OUTAGES CAUSED BY VEHICLES Dark Pole on Outside of Curve Rain Clearance Too Low Heavy Traffic Wet Surface Pole Too Close to Road ENVIRONMENT METHODS How should the Haasville team determine which root cause to attack first? 8 The Haasville team next constructs the Pareto Chart in Exhibit 6 on the next page. The Pareto Chart displays root causes of vehicle-related outages in decreasing order of their contribution to SUI. Although “Poles on Outside of Curve” is not necessarily #1 in terms of number of occurrences, it is #1 in terms of its contribution to SUI. So, the Haasville team chooses to attack the root cause “Poles on Outside of Curve”. 9 EXHIBIT 6: Pareto Chart for Root Causes of Vehicle-Related Outages 200,000 100% 190,000 180,000 90% 160,000 160,000 80% 140,000 70% Contribution to SUI 120,000 115,000 60% Contribution to SUI 100,000 50% Percent of Total 80,000 40% 60,000 30% 50,000 40,000 20% 20,000 15,000 10% 10,000 0 0% Pole on Outside Clearance Too Driver Fell Pole Too Close Construction Other of Curve Low Asleep To Road Equipment Root Cause 10 STEP 3: Identify alternative countermeasures that will mitigate the root cause, and select one of these countermeasures to implement. The Haasville team uses BRAINSTORMING to develop the following list of 11 alternative countermeasures: Relocate pole by moving it further back. Install barricade (doughnut) around pole. Move wires underground. Relocate pole to inside of curve. Install caution lights. Reduce speed limit. Install speed bumps on curve of road. Install reflectors on curbs at curve. Replace pole with break-away pole. Increase span between poles (i.e., decrease number of poles). 11 Replace pole with stronger pole. The Haasville team uses MULTIVOTING to reduce the entire list of 11 countermeasures to 3-5 countermeasures. (On the first vote, each team member can vote for any number of countermeasures. In any subsequent vote, each team member can vote for at most one-half the remaining countermeasures. The vote continues until the list of countermeasures is reduced to 3-5 items.) Using Multivoting, the Haasville team reduces the list of countermeasures to: A. Relocate pole by moving it further back. B. Relocate pole to inside of curve. C. Install speed bumps on curve of road. D. Install reflectors on curbs at curve. To evaluate the four alternatives, the Haasville team selects the following criteria: 1. Cost to Implement 2. Percentage of Problem Solved 12 3. External Assistance Required To assist in the selection of the “best” of the four remaining countermeasures, the Haasville team uses a technique for choosing the “best” from a set of options that have each been evaluated against a set of criteria. (One such technique is Prioritization Matrices.) After using Prioritization Matrices or a similar technique, the Haasville team selects the following countermeasure: Relocate poles currently located on outside of curve to inside of curve. How should the Haasville team decide the order in which the poles should be relocated? 13 STEP 4: If necessary, repeat Step 2 and/or Step 3, this time attacking another root cause of vehicle-related outages in specific or outages in general. STEP 5: Prevent the problem and its root cause from recurring. To prevent future problems with vehicle-related outages, the Haasville team designs a new process for locating new poles. Furthermore, the Haasville team takes steps to ensure that this new process is transferred to all of PE&G’s districts. STEP 6: Summarize lessons learned. The Haasville team answers the questions: What did we do well? What did we not so well? 14 What could we do differently?
Pages to are hidden for
"PE GS tory"Please download to view full document