20150812AVU to Staff 50 Attachment J.pdf

MAC 206 – Vegetation Management
Background
The Avista Vegetation Management program maintains the distribution and transmission
systems and strives to maintain a clearance between power lines and trees and other vegetation. The Vegetation Management program provides safety clearances for the public and also reduces customer outages caused by power lines coming into contact with trees. The Avista Electric
Distribution System runs for 7,777 overhead circuit miles in Washington, Idaho, and Montana.
The Transmission System includes 1,675 circuit miles of 115 kV Transmission Lines and 984
circuit miles of 230 kV Transmission Lines in Washington, Idaho, and Montana. The Gas System High Pressure Lines includes 291 miles. This report discusses strategic plans for Vegetation Management of the Electrical Distribution system, since vegetation management of
the Gas High Pressure Lines and Transmission System are driven by federal requirements.
Avista analyzed four cases to help determine an optimal approach for managing vegetation in the
electric distribution system.
1. The first case is called the No Action Case. In this case, Avista would only respond to outages and events (events are defined as a report of a tree or vegetation problem that
does not always result in an outage) as they occur. The No Action Case represents the
base case against which all other alternatives are compared. Avista compared this base
case against current practices and also examined two other alternatives. 2. The next case represents our current vegetation management practices (as of 2009) of trimming many feeders on either a 7 or 8 year cycle. This case is called the Current Case.
3. The first alternative represents trimming all Distribution Feeders on a straight five year
cycle and identifies it as the Five Year Case.
4. The second alternative came from optimizing the trim interval using our Asset Management software. This method optimized an interval for each individual feeder using the failure curves and is called the Optimized Case.
Based on risks associated with “risk trees” not previously identified for separate analysis, we
have added new scope for Vegetation Management to address codominant (split top) trees.
These codominant trees present a unique hazard because these split tree tops tend to fail at a significantly higher rate than normal tree tops.
Vegetation Management Program Trends and Analysis
Figure 1, Figure 2, and Figure 3 show the pattern of tree-related events tracked by Avista’s Outage Management Tool (OMT) for the last several years. Figure 4 shows how the current
trends indicate tree-related events might occur through 2016. Figure 1 shows the past six years
trends in the Outage Management Tool for Tree Fell, Tree Growth, and Tree Weather Events and
Figure 2 shows just the number of these events that actually cause an outage or partial outage. Figure 3 expands Figure 1 to show the number of events by quarter and year. Figure 4 shows an extrapolation of the current trends into the year 2016 for tree related OMT Events and the
extrapolation of the outages and partial outages. Comparing the curves on Figure 4, we see that
the number of OMT Events due to vegetation issues continues to grow but the impact from these
events has reduced. The number of OMT Events due to tree growth shows a rising trend but the
Staff_PR_050 Attachment J Page 1 of 24
outages from these events are actually trending down. Tree Fell events are also increasing at a
faster pace than the outages caused by tree falls. The trends in the average increment added to
SAIFI for Tree Fell, Tree Growth, and Tree Weather shown in Figure 5 supports the desired goal
that impacts from these types of outages has been improving.
The customer impacts based on the past three years are shown in Table 1 and Figure 6. The
average outage duration information used in Figure 6 does not go back any further than 2008, so
confidence to identify trends is questionable. The average number of affected customers has
remained relatively flat or declining for the past few years despite an increase in the number of
events.
In 2010, we trimmed 1,257 miles out of 7,800 circuit miles for a clearing cycle time of 6.2 years.
The approved amount for Washington state of $4.025 million provides sufficient funding to
support a 5.3 year clearing cycle time.
Overall, the data indicates the number of vegetation related events continue to increase.
However, the impact of vegetation related problems has improved based on the past reductions of the cycle times, so the severity of each event has diminished.
Vegetation Management Model Results
Sustained action is required to firmly establish a downward trend in the number of vegetation
related events. Based on our current analysis, the Optimized Case represents the best approach
to managing the system. The Optimized Case calls for tailoring the risks and program to each
feeder. The analysis indicates an average cycle time for each feeder of 4 years based on the
results shown in Table 2. Previous modeling indicated a cycle time of 5 years across the board for all feeders, but the updated model indicates better performance of the system in the Optimized Case and, therefore, we believe this justifies a shorter cycle time. Table 3 and Table 4
show the difference in system performance where the Optimized Cycle times will further reduce
vegetation events by an additional 380 events on average per year.
The data behind the failure rates used in the models comes from information gathered during past years’ work and failures. Failures are defined as a component or system did not properly perform a function and becomes a functional failure or simply a failure. Failure rate is defined as
the number of failures over a period of time and total population. The number of trees removed,
trees trimmed, and brush removed along with the tree-related failures documented in the OMT
were used to create the failure curves represented in by a combined curve in Figure 7. Examining the failure rate curve shown in Figure 7, we see that the failure rates due to weather, snow and ice, wind, tree growing into lines and trees falling into lines increases at a steady rate
in region 1. Region 1 of Figure 7 represents the current practices and its associated failures and
indicates that our system will reach a 53% failure percentage at about 10 years. The
corresponding five year trim cycle failure percentage is 22% and the four year trim cycle is at a 16% failure percentage. Region 2 of Figure 7 was caused by a lack of data and falsely implies the failure rate in this region is unchanged. Region 2 is an absence of information and is not
used. Region 3 of Figure 7 represents the failure rates associated with our old method of
trimming at 4 year intervals for predominantly deciduous or urban feeders and an 8 year interval
for predominantly conifer or rural feeders which are higher.
One effect not addressed by our analysis above come from the increased amount of work required to clear the distribution lines after delaying clearing for many years. Figure 8 shows
Staff_PR_050 Attachment J Page 2 of 24
how the amount of work grows over time. The first five years after clearing the line, the amount
of work remains relatively flat, but after 5 years, the work load grows exponentially. Figure 8 is
based on actual data collected that documented the number of trees trimmed, trees removed, and
amount of brush cleared by feeder and plotted against how many years have passed since the last time the line was cleared. Even after a line is cleared, some vegetation will quickly encroach on
the line because the removed vegetation held it back in the past.
Table 5 and Table 6 show the financial analysis between Base Case, our current practices, Five
Year Cycle Case, and the Optimized Case. Table 5 shows the impacts to our customers if we
ignore all possible consequence costs and customer outage costs. Table 6 shows the impacts to our customers when we consider all potential costs to the company and the customer outage
costs. To best determine the value of the program to our customers, we selected Table 6 to use
and maximize the value. The Optimized Case provides our customers the best value over time
by avoiding outage costs and other failure costs.
Our analysis clearly shows that a pro-active maintenance program is necessary to provide the best value to our customers. The Optimized Case (see Table 2) shows an optimized trimming
time of 4 years for most feeders with many feeders trimmed on a 3 or 5 year cycle. The
Optimized Case also provides the lowest predicted number of vegetation related events and
outages (see Table 4). The force behind the analysis that drives the Optimized Case to an
average trim cycle of four years is the risk associated with fires and property damage.
The required budgets for 2011 through 2015 are shown in Table 7 for Washington and Table 8
for Idaho (combined numbers for both states are shown in Table 9). These estimates are based
on model predictions and on the feeders scheduled for work over the next 5 years. Based on the
proposed budget, Avista will need an additional $4.4 million over the test year of 2010
($3,144,720 for Washington and $1,178,410 for Idaho as shown in Table 11). Figure 9 plots the cumulative costs for each alternative over time and demonstrates how the value will be realized.
Future savings estimates will be realized only after the work is completed and generally will not
be seen until the year following the actual work. This means that the actual savings for work
performed in 2011 will not be realized until 2012. The estimates for savings in the future are
based on Table 10. The assumption is that all of the events over 5 years in will be avoided on a 5 year trim cycle and save ~ $1,000 per event. This assumption is very optimistic and the actual
number will be less because some events will still occur within a 5 year trim cycle as we have
seen for all of the feeders discussed above. Next, the total savings cannot be realized until all of
the feeders have been trimmed on a 5 year cycle, so the savings for 2012 is 1/5 of the overall
savings and is shown in Table 10 as $80,000 for Idaho and $154,800 for Washington in 2012. The four year cycle will save an additional $342,000 after the first year of change (assumed to be
2013) as shown in Table 5.
Managing Vegetation Risk
All trees have the potential to fail, and any tree failure has the potential to contact electric power
lines. Trees or their branches may fall into power lines or they can be carried into contact from a
distance beyond the reach of gravity’s pull in high wind situations. Currently, field reviews have
shown that the majority of tree related outages are due to tree failures and most of these trees are outside of the right of way. It has also been found that tree failures increase rapidly at wind
Staff_PR_050 Attachment J Page 3 of 24
speeds over 60 mph. Avista seeks to drive down outages due to tree failures through an
enhanced risk tree removal program.
A risk tree is defined as dead, diseased or dying tree, however, many trees that fail have no
visible fault indicators which make it difficult to justify as removals. Risk assessment does not mean just finding dead trees and branches, as any defects such as internal decay are hidden and
require a higher level of inspection to identify.
Studies show common annual tree mortality to be 1% to 3% and there is an ongoing residual
population of risk trees. Tree mortality rates will depend on tree species and conditions.
Based on increased risk, Avista is modifying our current risk tree program. Codominant stemmed trees will be removed, or structurally pruned in a fashion where a dominant leader is
developed, which is usually the largest stem. Addressing the codominant stemmed trees is new
to our current program and is discussed in more detail in Appendix A. Avista will also revise the
current risk tree inspection process to inspect and remove trees more frequently to insure right of
ways are reviewed every two years. We will also focus on problem tree species. The additional codominant stemmed tree scope and revised risk tree inspections drives some of the increased
costs planned for 2012.
Summary
In summary, Avista plans to change the Vegetation Management Program to an optimized cycle
based on the individual feeders for an average cycle time of four years. We will also go to a two
year cycle on inspecting and removing risk trees on all circuits. The Optimized Case provides
the best return to the customers by avoiding significantly more outages than the other alternatives and avoiding an increasing risk associated with vegetation caused fires. In order to get the Vegetation Management Program into the correct cycle, Avista will need an additional $4.4
million above the spending in 2010 in order to keep future rates lower.

Staff_PR_050 Attachment J Page 4 of 24

Figure 1, Vegetation Related OMT Events by Year
0
100
200
300
400
500
600
700
800
900
1000
Nu
m
b
e
r

o
f

O
M
T

E
v
e
n
t
s

Year
Tree Fell Tree Growth Weather
Staff_PR_050 Attachment J Page 5 of 24
Figure 2, Vegetation Management Related Outages and Partial Outages by Year
0
100
200
300
400
500
600
700
Nu
m
b
e
r

o
f

O
M
T

O
u
t
a
g
e
s

a
n
d

P
a
r
t
i
a
l

Ou
t
a
g
e
s

Year
Tree Fell Tree Growth Weather
Staff_PR_050 Attachment J Page 6 of 24
Figure 3, Vegetation Management OMT Events for Tree Fell and Tree Growth by Quarter
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
2005 2006 2007 2008 2009 2010
Nu
m
b
e
r

o
f

O
M
T

E
v
e
n
t
s
Year and Quarter
Tree Fell Tree Growth
Staff_PR_050 Attachment J Page 7 of 24
Figure 4, Vegetation Management OMT Events and Outages
0
200
400
600
800
1000
1200
1400
1600
0
200
400
600
800
1000
1200
1400
1600
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Nu
m
b
e
r

o
f

O
u
t
a
g
e
s

a
n
d

P
a
r
t
i
a
l

O
u
t
a
g
e
s
Nu
m
b
e
r

o
f

O
M
T

E
v
e
n
t
s
Year
Tree Fell Events Tree Growth Events Tree Weather Events
Tree Fell Outages Tree Growth Outages Tree Weather Outages
Staff_PR_050 Attachment J Page 8 of 24
Figure 5, Average Incremental Impact to SAIFI by Tree Fell, Tree Growth, and Tree Weather Outages
0
0.0005
0.001
0.0015
0.002
0.0025
2005 2006 2007 2008 2009 2010
Av
e
r
a
g
e

I
m
p
a
c
t

t
o

S
A
I
F
I

p
e
r

O
u
t
a
g
e
Year
Tree Fell Tree Growth Weather
Staff_PR_050 Attachment J Page 9 of 24
Figure 6, Trends in Average Number of Affected Customers and Average Outage Durations

Table 1, 3 Year Average of Customer Impacts from Vegetation Related Outages
Subreason
Tree Fell 2.294 104.030
2.757 23.110
3.208 124.438

0
0.5
1
1.5
2
2.5
3
3.5
4
0
50
100
150
200
250
2004 2005 2006 2007 2008 2009 2010 2011
Av
e
r
a
g
e

O
u
t
a
g
e

D
u
r
a
t
i
o
n

(
h
r
s
)
Av
e
r
a
g
e

N
u
m
b
e
r

o
f

A
f
f
e
c
t
e
d

C
u
s
t
o
m
e
r
s
Year
Tree Fell -Average Number of Impacted Customers Tree Growth -Average Number of Impacted Customers
Tree Weather -Average Number of Impacted Customers Tree Fell -Average Outage Duration
Tree Growth -Average Outage Duration Tree Weather -Average Outage Duration
Staff_PR_050 Attachment J Page 10 of 24
Table 2, Optimized Vegetation Management Cycle Times Optimized Interval (Years) Number of Feeders
1 1
2 1
3 24
4 271
5 21

Table 3, OMT Data and Projection Comparisons for Current Case and 5 Year Cycle Case
OMT Events
420 309 440 1,169
Average - Current 330 789 774 1,893
Average - 5 Year 212 344 164 720
Current Case and 5
Year Trim Cycle 118 445 610 1,173

Staff_PR_050 Attachment J Page 11 of 24
Table 4, OMT Data and Projection Comparisons for Current Case and Optimized Case
OMT Events
420 309 440 1,169
Average - Current 330 789 774 1,893
Average - 53 225 62 340
Current Case and 277 564 712 1,553

Table 5, Projected O&M Offsets for changing from a 5 Year to 4 year Clearing Cycle
Year
Tree Fell
Outage Cost
per Event
Tree Growth
Outage Cost
per Event
Tree
Weather
Outage Cost
per Event
Value if
changed
from 5 year
to 4 year
$502.65 $2,489.23 $163.91 $0
$517.73 $2,563.91 $168.83 $342,000
$533.27 $2,640.83 $173.89 $704,000
$549.26 $2,720.05 $179.11 $1,088,000
$565.74 $2,801.65 $184.48 $1,495,000

Table 6, Financial Analysis excluding Effects and Consequence Costs
Title Customer Internal
Rate of Return
Levelized Gr. Mar.
Requirement
No Action
Case 18.16% $4,823,333
Current Case 4.03% $5,813,678
5 Year Case 0.70% $7,225,887
Optimized
Case -2.79% $9,207,982
Staff_PR_050 Attachment J Page 12 of 24

Table 7, Financial Analysis including Effects and Consequence Costs
Title Customer Internal
Rate of Return
Levelized Gr. Mar.
Requirement
No Action Case -22.91% $157,778,097
Current Case 38.59% $38,076,500
5 Year 63.39% $19,207,108
Optimized Case 75.96% $14,790,077

Staff_PR_050 Attachment J Page 13 of 24
Table 8, Washington’s Distribution Vegetation Management Projected Budget
WASHINGTON
2011 2012 2013 2014 2015
Circuit Trimming $3,917,679 $5,154,805 $5,273,366 $5,394,653 $5,518,730
Unplanned Trimming $200,000 $175,000 $150,000 $125,000 $125,000
Sub Total $4,117,679 $5,329,805 $5,423,366 $5,519,653 $5,643,730
HPGPL Reclearing $50,000 $100,000 $50,000 $50,000 $50,000
Avista FTE (1 - Tim) $70,000 $72,100 $75,000 $77,000 $80,000
Enhanced Risk Tree / CoDom N/A $313,816 $201,568 $315,315 $108,674
Contract Administration $334,000 $354,040 $375,000 $398,000 $422,000
Grand total WASHINGTON $4,571,679 $6,169,761 $6,124,934 $6,359,968 $6,304,404

Table 9, Idaho’s Distribution Vegetation Management Projected Budget
IDAHO
2011 2012 2013 2014 2015
Circuit Trimming $1,533,350 $2,780,532 $2,844,484 $2,909,907 $2,976,835
Unplanned Trimming $100,000 $100,000 $100,000 $100,000 $100,000
Sub Total $1,633,350 $2,880,532 $2,944,484 $3,009,907 $3,076,835
HPGPL Reclearing $100,000 $150,000 $50,000 $50,000 $50,000
Avista FTE (1 - Ernie) $70,000 $72,100 $75,000 $77,000 $80,000
Enhanced Risk Tree / CoDom N/A $104,845 $171,299 $149,849 $171,123
Contract Administration $170,000 $180,000 $191,000 $202,000 $215,000
Grand total IDAHO $1,973,350 $3,387,477 $3,431,783 $3,488,756 $3,592,958

Table 10, Combined Washington and Idaho Vegetation Management Projected Budget
GRAND TOTAL D51
2011 2012 2013 2014 2015
Circuit Trimming $5,451,029 $7,935,337 $8,117,850 $8,304,560 $8,495,565
Unplanned Trimming $300,000 $275,000 $250,000 $225,000 $225,000
Sub Total $5,751,029 $8,210,337 $8,367,850 $8,529,560 $8,720,565
HPGPL Reclearing $150,000 $250,000 $100,000 $100,000 $100,000
Avista FTE (2) $140,000 $144,200 $150,000 $154,000 $160,000
Enhanced Risk Tree / CoDom N/A $418,661 $372,867 $465,164 $279,797
Contract Administration $504,000 $534,040 $566,000 $600,000 $637,000
Grand total $6,545,029 $9,557,238 $9,556,717 $9,848,724 $9,897,362
Staff_PR_050 Attachment J Page 14 of 24
Figure 7, Unreliability due to Snow/Ice/Wind/Weather/Tree Growth/Tree Fell Combined based on Time since last Trimmed
Veg Man Combined Cumulative Probability
24.33 441.4 8008 1.453E+05
Time
0.1
0.2
0.3
0.5
1
2
3
5
10
20
30
50
70
90
99
99.9
Un
r
e
l
i
a
b
i
l
i
t
y

(
%
)
Eta estimator
P0: 0%
B20: 4.065E+04
B15: 3.334E+04
B10: 2.543E+04
ε: 0.02255
γ3: 6.47Ε+04
β3: 1.548
η3: 6.577Ε+04
γ2: 4.574Ε+04
β2: 1
η2: 2.514Ε+06
γ1: 0
β1: 1.6
η1: 1.038Ε+05
Median rank
Phased tri-Weibull
so not included in
analysis
Staff_PR_050 Attachment J Page 15 of 24
Figure 8, Probability a 150’ Feeder Segment will require Vegetation Management based on Time since Last Trimmed
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
120.00%
0 2 4 6 8 10 12 14 16
Pr
o
b
a
b
i
l
i
t
y

a

L
i
n
e

S
e
g
m
e
n
t

w
i
l
l

R
e
q
u
i
r
e

W
o
r
k
Years since Last Vegetation Management
Probability a Feeder Segment will require Vegetation Management
Work
Staff_PR_050 Attachment J Page 16 of 24

Table 11, Savings Offset Estimate
Management
Related OMT
Events
Number of
Potentially
Avoided
Total Savings
after 10 Years
Offset

Table 12, Distribution Pro Forma Increment
Year WA Elec ID Elec
2010 Actual $3,144,720 $1,178,410
2011 Planned $4,521,679 $1,873,350
2012 Planned $6,069,761 $3,237,477
2012 Offset -$154,800 -$80,000
$2,770,241 $1,979,067
Staff_PR_050 Attachment J Page 17 of 24

Staff_PR_050 Attachment J Page 18 of 24

0
200
400
600
800
1000
1200
1400
1600
Cu
m
u
l
a
t
i
v
e

C
o
s
t

P
r
o
j
e
c
t
i
o
n
s
Mi
l
l
i
o
n
s

Year
Staff_PR_050 Attachment J Page 19 of 24

Figure 9, Cumulative Cost Comparison Curve
0
200
400
600
800
1000
1200
1400
1600
Cu
m
u
l
a
t
i
v
e

C
o
s
t

P
r
o
j
e
c
t
i
o
n
s
Mi
l
l
i
o
n
s

Year
Staff_PR_050 Attachment J Page 20 of 24
Appendix A - Enhanced risk tree and development of co-dominant stem removal programs

Staff_PR_050 Attachment J Page 21 of 24
2011 Rate Case information
Enhanced risk tree and development of co-dominant stem removal programs

BACKGROUND – Codominant Stemmed Trees

Coniferous trees with good structure are characterized by a single dominant leader, strong branch unions without bark inclusion,
whereas, trees with compromised structure, such as ones with a co-dominant stem, have more potential to fail in severe weather
events. Many studies have found that trees fail due to structural issues like codominant stems and bark inclusions.

Codominant stems originate from the same point on the tree. Included bark is bark pinched between two stems creating a weak
union, and stems with a tight ‘V’ shaped union are often accompanied by included bark. This union is weak because the bark
inclusion prevents any physical connection between the two stems. A co-dominant stem will split from a tree because of a weak
branch connection and included bark.

Most codominant stemmed trees are healthy, intact trees which can be negatively impacted by extreme wind, snow and ice loading.
Codominant stems that fail can make a bridge between phases resulting in sparking and arching, which could result in a devastating
fire and/or an overall outage of the line.

There will be a need to employ additional qualified/certified tree assessors and software developed to address the codominant stem
population throughout Avista’s service territory.

BACKGROUND – Enhancement of Risk Tree Identification and Removal Program

All trees have the potential to fail, and any tree failure has the potential to cause property damage, injury or death. Currently, field
reviews have shown that the majority of tree-related outages are due to tree failures and most of these trees are outside of the right
of way. Avista seeks to drive down outages due to tree failures through an enhanced risk tree removal program.

A risk tree is defined as dead, diseased or dying tree, however, a many of trees that fail have no visible fault indicators which make it
difficult to justify as removals. Risk assessment does not mean just finding dead trees and branches, as any defects such as internal
decay are hidden and require a higher level of inspection to identify.

Staff_PR_050 Attachment J Page 22 of 24
Studies show common annual tree mortality to be 1% to 3% and there is an ongoing residual population of risk trees. Tree mortality
rates will depend on tree species and conditions. It has been found that tree failures increase rapidly at wind speeds over 60 mph.

Risk is directly related to the number of trees within striking distance of the line. All trees are susceptible to failure and all trees
capable of striking the line represent liability. The risk can be mitigated by decreasing the number of trees capable of striking the
line.

A risk assessment program allows for prudent use of ratepayer dollars because it clearly delineates action thresholds based on risk
tolerance, and ensures that money is being spent in areas that provide the biggest reliability return and lowers liability, both for the
least cost.

Avista plans to follow up with risk tree evaluation 2 years after the cycle trim is completed on specific circuits which dictate needing
the work performed.

There will be a need to employee additional qualified/certified tree risk assessors and software developed to address this
enhancement and to identify the risk tree population throughout Avista’s service territory.

IN CONCLUSION –

Tree related outages are an expression of the tree liability, therefore resulting in the utility’s overall liability associated with tree
failures.

DEVELOPMENT OF CODOMINANT STEM REMOVAL PROGRAM - Codominant stemmed trees should either be removed, or
structurally pruned in a fashion where a dominant leader is developed, which is usually the largest stem.

ENHANCEMENT OF RISK TREE REMOVAL PROGRAM - Avista would like to revise the current risk tree inspection process to a cycle
being performed on a more frequent basis to insure right of ways are reviewed two years after normal cyclic trimming has been
completed. We will also focus on problem tree species.

DEVELOPMENT OF AN ELECTRONIC TREE INVENTORY SYSTEM – Avista needs to review, inventory and assess its tree population.
We are proposing the development of an electronic data gathering system to capture that information. This data will be our base in
Staff_PR_050 Attachment J Page 23 of 24
the development, prioritization and scheduling associated with the annual integrated vegetation management work plan(s). The
data will also be utilized for reporting purposes.

We will utilize GIS/AFM to review past tree related OMT events as an additional tool to help develop and prioritize the work plan(s).

ADDITIONAL FTE REQUIREMENTS (Per our conversation 3/3/11) –

• 1 Program Manager
• 2 pre-auditors / work planners
• 1 technician (focus on AFM/GIS/Electronic data gathering information and other duties)
• At least 2 each 3-man tree crews

Staff_PR_050 Attachment J Page 24 of 24