Development of a Methodology to Evaluate Waste and Recycling Rates

Debra L. KantnerBryan Staley, PhD, PE

Our History and Mission

• Founded in 1992 as a 501(c)3 charity

• Mission: To direct research studies & educational initiatives for sustainable waste management practices via:– Research grants– Scholarships/Internships– General education

Key Programs

• RESEARCH Nearly $11 million in projects that help shape policy, develop sustainable practices, bring value, and direct the future of solid waste management

• SCHOLARSHIPS Educating the next generation of solid waste researchers and technical personnel – nearly 50 scholars thus far, totaling almost $0.75 million

• CONTINUING EDUCATION Informing policymakers, practitioners, and others regarding cutting edge research and solid waste management practices via meetings and online continuing education

• INTERNAL RESEARCH Conduct state of practice research and trend analyses to benefit to solid waste industry, while providing research experience for talented college undergraduates.

Strategic Focus Areas

EREFs strategic plan addresses all areas of the integrated waste management infrastructure.

1) Landfills

2) Equipment/Safety

3) Transport/Collection

4) Policy/Economics

5) Recycling/Waste Minimization (includes packaging)

6) Combustion/WTE

7) Conversion Technologies (includes composting)

8) Life Cycle Analysis/Inventory

Development of Waste and Recycling Rate Methodology

• Current methods (e.g. US EPA) result in:– Estimation of recycling rates on a citywide basis– ‘Blanket-style’ approach to management and

incentives• Such methods are not specific enough to

assist cities/haulers in identifying rate trends.

• Recent improved tracking techniques allow for a data-driven approach to provide better information to the end user (e.g. city, hauler)

Background

Overview

Methodology developed by EREF provides two sets of analyses:

1) Rates Analysis for specific areas within a city using available technology or operations data (e.g. RFID, ‘clicker’/driver collected, etc)

2) Correlation Analysis that indicates why particular areas may have depressed or elevated rates

Can be applied to BOTH waste and recycling data

Basic Methodology

1) Obtain collection data– Weekly participation based on collection events– Mass per route based on weigh tickets

2) Quantify rates geospatially based on:– Truck route– Census tract/block– Individual residence– Custom boundaries (neighborhood, street, etc.)

3) Determine high/low participation areas – Set-out rates– Average mass per home per route

• High or Low participation are typically assigned in reference to the city average

4) Analyze fraction of service area in high/low participation areas– How much of area affected, # carts, # homes, etc– How far below average for each area

Basic Methodology(continued)

5) Determine why certain areas may have high/low participation based on:– Demographics– Housing– Other characteristics

6) Develop recommendationsCan include:– Target incentive programs based on demographics– Determining size of targeted incentive groups

Basic Methodology(continued)

Results

Assessing Rates via a Data Driven Approach

Colors indicate individual truck

service area

Each icon is an individual pickup

RFID data for recycling carts from

March 1, 2011to

Feb 24, 2012

Rates Analysis

Rate Definitions

Set-Out Rate:

The percentage of set-out opportunities that were utilized by program participants during a defined period of time.

Mass Rate:

The amount of recyclables set at the curb, based on route average. This is determined using scale tickets

Set-Out Rates(based on Census tracts)

• Citywide weekly set-out rate: 63.6%– 41 tracts– 62,440 carts– 4.2M collection events– Varied 43% to 78%

• 35 point difference

Set-Out Rate

Observed Cart Mass

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 440

10000

20000

30000

40000

50000

60000

Distribution of Mass per cart

lb/cart

# o

f O

bs

erv

ati

on

s

Average 28.1 lb/cart

Min 11.5 lb/cart

Max 49.6 lb/cart

Mass Rates(based on block group)

• Citywide average mass rate: 400 lb/HH-yr– 82 block groups– 62,440 carts– 4.2M collection events– Varied 231 to 639 lb/HH-yr

• 408 pound difference

Mass Rate (lb/HH-yr)

231 – 334334 – 435435 – 537537 – 640

Defining High/Low

• The citywide average weekly participation rate and standard deviation are used to define “high” and “low” groups

Defining “High” and “Low” Groups

Citywide Average 63%

Standard Deviation 9%

“High” > 72%

“Low” < 54%

Low HighAverage

High/Low Comparison(based on census tracts)

• HIGH: >72%

set-out rate– 6 tracts– 9,615 carts– 16% of carts

• LOW: < 54%

set-out rate– 8 tracts– 9,163 carts– 15% of carts

Correlation Analysis

Participation v. mass

200 250 300 350 400 450 500 550 600 650 7000.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

Household generation rate (lb/HH-yr)

Mass rate, lb/HH-year

Set

ou

t ra

te

0 50 100 150 200 250 3000.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

Per capita generation rate (lb/person-yr)

Per capita mass rate, lb/person-year

Set

ou

t ra

teParticipation v. mass

• Examined 96 Census Bureau variables to find possible relationships to participation rate.

• Performed statistical analysis to determine which variables were significant.

• Result: 5 primary variables were most important and correlated to participation

Correlation Variables

Set-Out Rate

Correlation Plots

0 10 20 30 40 50 60 70 80 90 1000.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

R² = 0.644069246623594

Household Income

% of Households earning $75,000/year or more

Wee

kly

Set

-Ou

t R

ate

0 10 20 30 40 50 60 70 80 90 1000.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

R² = 0.397063558406296

Education Level

% Population with Bachelor's Degree or Higher

Wee

kly

Set

-Ou

t R

ate

Correlation Plots(continued)

Correlation Plots(continued)

50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,0000.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

R² = 0.578475847306635

Home Value

Median Home Value ($)

Wee

kly

Set

-Ou

t R

ate

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

R² = 0.572940249459981

Multi-Resident Households

% of Households with multiple residents

Wee

kly

Set

-Ou

t R

ate

Correlation Plots(continued)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

R² = 0.474062021960916

Owner Occupancy

% of Households that are owner-occupied

Wee

kly

Set

-Ou

t R

ate

Correlation Plots(continued)

Mass Rate

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0

100

200

300

400

500

600

700

R² = 0.510442286066574

Household Income

% of HH earning $75,000/year or more

Mas

s R

ate,

lb

/HH

-yr

Correlation Plots

100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 550,0000

100

200

300

400

500

600

700

R² = 0.467447063153152

Home Value

Median Home Value (US $)

Mas

s ra

te,

lb/H

H-y

ear

Correlation Plots

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0

100

200

300

400

500

600

700

R² = 0.420811614740489

Education Level

Percent of People with a bachelor’s degree or higher

Mas

s ra

te,

lb/H

H-y

ear

Correlation Plots

Household Income

Education Level

Home Value Household Size

Owner Occupancy

Set-Out% YES YES YES YES YES

MassLb/HH-yr YES YES YES NO NO

Correlation Summary

Townhouse Analysis

• Sampled within city to test observed participation differences between single family homes and townhomes.

– 7 tracts– 2% to 57% townhouses

Townhome v. Single Family

Townhome v. Single Family(continued)

TownhomesSingle Family

Homes

Set-Out RateAverage

45% 64%

Set-Out RateRange

38% - 53% 55% - 77%

Mass RateAverage

370 lb/HH-yr

507lb/HH-yr

Mass RateSt. Dev.

177lb/HH-yr

207lb/HH-yr

Average Set-Out RateTownhomes v. Single Family

• Tracts for townhouse comparison include high and low:– Set-out rate– Income– Education– Owner Occupancy

• Both downtown units and neighborhood complexes

Methodology Benefits

• Identify key differences within the low tracts

• Tailor programs to tract demographics– Incentivize high and

low income areas differently (A and B)

– Examine bin size and collection frequency in C

CensusTract

Low Education

Low Income

Low Home Value

Individual Resident

RenterOccupied Townhouse

A X X

B X X X

C X X

D X

Recommendations

– Townhouses may benefit from education/awareness

Benefits

1) Targeted spending of program budget

2) Allows for more effective implementation of awareness and incentive programs

3) Provides a means to track performance over time that is coupled with demographics and program data (e.g. before and after analyses)

4) Data provides opportunity for further analysis

Further Analysis

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1f(x) = 1.22421952799121 x

Set-out rate v. participation rate

Observed Set-Out rate

Ob

serv

ed

Par

tici

pat

ion

Rat

e

Participation Estimate

0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.81

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

Set-out rate v. Participation multiplier

Observed Set-Out Rate

Rea

l M

ult

ipli

erParticipation Estimate

Thank You