Measuring the economic and societal value of reliability/resilience investments: case studies of islanded communities

Figures & data

Figure 1. Investments to reduce the risks of power interruptions without and with resilience-based planning and informed power interruption cost estimates.

Figure 2. Diagram of the steps of eliciting informed customer interruption cost (CIC) estimates and constructing duration-dependent customer damage functions.

Table 1. Power interruption cost and saving components that need to be elicited from each customer segment. Asterisks indicate the costs or saving components that are not captured in the traditional CIC surveys.

Customer Class	Direct Cost and Saving Components
Residential	WTP for hypothetical backup service
Small and medium non-residential (SMNR)	Additional labor costs to make up for lost production Lost revenue Damage to raw materials/inventory Damage to equipment Additional labor costs to address outage-related issues Other tangible costs, including backup generation cost Reduced personnel costs Savings due to an outage Losses covered by insurance*
Large non-residential (LNR)	Cost to operate backup generation Additional labor costs to make up for lost production Lost revenue Damage to raw materials/inventory Damage to equipment Additional labor costs to address outage-related issues Other tangible costs Reduced personnel costs Savings due to an outage Losses covered by insurance*
Public	Additional costs associated with addressing the effects of a power outage Percentage increase in response time, number of injuries, and number of deaths during the outage*

Figure 3. Illustration of elicitation design for residential (top), SMNR/LNR (middle), and public customers (bottom).

Table 2. Summary of the initiating events and power interruption scenarios.

Duration	Region 1: Gulf of Alaska	Region 2: Parts of Southeast and Interior Alaska	Region 3: U.S. Territory in South Pacific
A six-hour-long power interruption	Equipment failure Affected a few neighboring electricity customers Winter	Equipment failure Affected a few neighboring electricity customers Winter	Equipment failure Affected a few neighboring electricity customers Summer
A two-day-long power interruption	Volcanic activity Entire census area Winter	A tsunami with a landslide (Southeast Alaska) or heavy snowfall (Interior Alaska) Entire census area Winter	A category 3 typhoon Entire territory Summer
A seven-day-long power interruption	Earthquake Entire Southcentral and Southeast Alaska Winter	A tsunami with a strong earthquake (Southeast Alaska) or a winter storm (Interior Alaska) Entire Southeast or Interior Alaska Winter	A tsunami with a strong earthquake Entire U.S. territory Summer

Figure 4. Distribution of the interruption cost estimates collected from surveys and the fitted quadratic CDFs for the part of Southeast and Interior Alaska.

Table 3. Duration-dependent customer damage functions (CDFs) for the direct power interruption costs.

Customer Class	Direct CDF: Gulf of Alaska	Direct CDF: Parts of Southeast and Interior Alaska	Direct CDF: U.S. Territory in South Pacific
Residential	($3.2+0.092\mathit{t}{)}^2$	$19+1.7\mathit{t}-0.0022t^2$	$(4.8+0.044\mathit{t}{)}^2$
SMNR	$(25+0.54\mathit{t}{)}^2$	$(56+0.78\mathit{t}{)}^2$	$(63+0.64\mathit{t}{)}^2$
LNR	$(110+2.1\mathit{t}{)}^2$	$(54+0.84\mathit{t}{)}^2$	$(210+3.4\mathit{t}{)}^2$
Public	$(79+1.6\mathit{t}{)}^2$	$7500+130\mathit{t}+1.4t^2$	$(75+3.2\mathit{t}{)}^2$

Table 4. Duration-dependent CDFs for the indirect power interruption costs.

Customer Class	Indirect CDF: Gulf of Alaska	Indirect CDF: Parts of Southeast and Interior Alaska	Indirect CDF: U.S. Territory in South Pacific
SMNR	$0.24(25+0.54\mathit{t}{)}^2$	$0.5(56+0.78\mathit{t}{)}^2$ to $2(56+0.78\mathit{t}{)}^2$	$0.5(63+0.64\mathit{t}{)}^2$ to $2(63+0.64\mathit{t}{)}^2$
LNR	$0.31(110+2.1\mathit{t}{)}^2$	$0.5(54+0.84\mathit{t}{)}^2$ to $2(54+0.84\mathit{t}{)}^2$	$0.5(210+3.4\mathit{t}{)}^2$ to $2(210+3.4\mathit{t}{)}^2$
Public	$0.29(79+1.6\mathit{t}{)}^2$	$0.5\left(7500+130\mathit{t}+1.4t^2\right)$ to $2\left(7500+130\mathit{t}+1.4t^2\right)$	$0.5(75+3.2\mathit{t}{)}^2$ to $2(75+3.2\mathit{t}{)}^2$

Table 5. Expected percent increase in morbidity and mortality provided by the critical private and public organizations in the study regions.

Outage scenario	Gulf of Alaska	Parts of Southeast and Interior Alaska	U.S. Territory in South Pacific
Two-day-long power interruption	30% increase in injuries 10% increase in deaths	10–15% increase in injuries	5–10% increase in injuries
Seven-day-long power interruption	60% increase in injuries 35% increase in deaths	20–25% increase in injuries 25% increase in deaths	40–50% increase in injuries

Table 6. Key inputs and assumptions used to determine the health/safety CDFs for the Gulf of Alaska.

Assumptions	Sources/Comments
Injury hospitalization rate is 115.4 per 10,000	Alaska Native Injury Atlas (Strarke, Stevens, et al., 2020)
Injury-related death rate is 182.4 per 100,000	Alaska Native Injury Atlas (Strayer et al., 2020)
Population of the study region is 2,829 in 2020	State of Alaska Department of Labor and Workforce Development (2020)
Value of statistical life is $12,100,000 (VSL)	U.S. Department of Transportation (2021)
Cost of non-fatal injury is $41,570 per patient	Peterson et al. (2021)
50% of the morbidity and mortality-related issues are associated with the power interruptions	The remaining 50% of morbidity- and mortality-related costs are attributed to the natural disaster (Sheridan, Zhang, Deng, et al., 2021)

Table 7. Summary of the comparison of the islanded communities’ direct and indirect power interruption cost estimates (calculated from the duration-dependent CDFs) and those elicited from San Francisco, California.

Customer Class	Gulf of Alaska	Parts of Southeast and Interior Alaska	U.S. Territory in South Pacific: Estimated cost of one-week-long power interruption ($/customer)	San Francisco: Estimated cost of one-week-long power interruption ($/customer)
Residential	$350	$240	$150	$180
SMNR	$16,600	$52,500 to 105,000	$44,000 to $87,000	$200,000 to $410,000
LNR	$281,000	$57,000 to 114,000	$920,000 to $1,800,000	$830,000 to $1,700,000
Public	$156,000	$103,000 to 207,000	$560,000 to $1,100,000	N/A

Table 8. Summary of the comparison of the islanded communities’ direct and indirect power interruption cost estimates and those elicited from the Village of Rockville Centre, New York.

Customer Class	Gulf of Alaska: Estimated cost of one-week-long power interruption ($/customer)	Parts of Southeast and Interior Alaska	U.S. Territory in South Pacific: Estimated cost of one-week-long power interruption ($/customer)	Village of Rockville Centre: Estimated total cost of one-week-long power interruption ($)
Residential	$1.0 M	$0.73 M	$0.45 M	$40 M
SMNR	$7.3 M	$24 M-47 M	$20 M-39 M
LNR	$14 M	$2.9 M–5.7 M	$46 M-92 M
Public	$4.1 M	$2.7 M–5.4 M	$15 M-29 M

Strayer, H., Blake, I., Stevens, I., & Provost, E. (2020). Alaska Native Injury Atlas. Third. Alaska Native Tribal Health Consortium Injury Prevention Program and Alaska Native Epidemiology Center. Available athttp://anthctoday.org/epicenter/publications/InjuryAtlas2020/2020_AlaskaNative_InjuryAtlas_FullReport.pdf.

 Google Scholar

U.S. Department of Transportation. (2021). Guidance on the treatment of the economic value of a statistical life (VSL) in U.S. Department of transportation analyses - 2021 update. https://www.transportation.gov/sites/dot.gov/files/2021-03/VSL%20Update%202021%20-%20Transmittal%20Memo.pdf

 Google Scholar

Peterson, C., Xu, L., & Florence, C. (2021). Average medical cost of fatal and non-fatal injuries by type in the USA. Injury Prevention, 27(1), 1–10.

 PubMed Web of Science ®Google Scholar

Sheridan, S. C., Zhang, W., Deng, X., & Lin, S. (2021). The individual and synergistic impacts of windstorms and power outages on injury ED visits in New York State. Science of the Total Environment, 797, 149199. https://doi.org/10.1016/j.scitotenv.2021.149199

 PubMed Web of Science ®Google Scholar

Data availability statement

The (anonymized) data supporting this study’s findings are available upon request from the authors.