ASHRAE Journal - December 2019 - 16

TECHNICAL FEATURE

Factors Influencing Persistence
We tested our data set for several correlation factors
that could potentially impact savings persistence. At
each site, we interviewed and gathered responses from
the facilities personnel on building operation and maintenance. Based on this data, we identified key metrics
against which to test correlations with energy savings
persistence. Some factors showed a clear correlation
with persistence, such as sites that have had a major
16

ASHRAE JOURNAL

ashrae.org

D E C E M B E R 2 0 19

TABLE 3 Overall persistence weighted by electric energy savings per measure.
PERIOD

PERSISTENCE STANDARD
(ELECTRIC ENERGY
ERROR
SAVINGS WEIGHTED)

90% CONFIDENCE LOW

90% CONFIDENCE HIGH

All

0.61

0.09

0.46

0.76

PY3

0.76

0.09

0.61

0.92

PY6

0.49

0.11

0.30

0.68

FIGURE 1 Persistence by measure category.

Filters
0.8

0.6
Persistence

in savings. From this analysis, persistence based
on measure count alone (not the weighted electric
energy savings) was found to be 58.1% (±8.6%). Sites
analyzed six years after RCx and three years after RCx
showed 63% (±15%) and 52.8% (±8.9%) persistence,
respectively.
More significantly, we found a 60.8% (± 8.9%) persistence in weighted electric energy savings across the
program (Table 3). The 2010/11 and 2013/14 data sets
showed 76.4% (± 9.1%) and 49% (± 11.2%) persistence
respectively.
We expected 2013/14 to have higher persistence
compared to the 2010/11 data set since RCx energy
savings will degrade over time. But our results
showed higher persistence for 2010/11 compared
to 2013/14. Our investigation into the 2014 data set
revealed that the low persistence was driven by two
very large commercial facilities that had very low persistence at the building level (0.09 and 0.34). Since
these two buildings contributed to the highest kWh
savings in the year 2014, it negatively influenced the
overall persistence for the year 2014.
A persistence higher than 50% reflects that a measure
is still within its useful life. We reviewed all measures
in each measure category for persistence by measure
type. The number of data points we were able to collect
varied by category. Some measures such as economizer
and OA control, and plant control optimization had 32
observed instances each, while a few other categories did
not have any available data points.
Filters and airflow adjustment measures showed the
highest persistence. Lighting controls, optimum start
for AHU's and thermostat settings had the lowest performance with zero persistence. The sample size for these
categories was very small, and we only calculated 90%
confidence interval for measure categories with more
than 14 instances.

0.4

System Schedule
Static
SAT
Pressure
Reset
Reset
Other
Economizer
And OA Control

Airflow Adjustment
Plant Sequencing
And Isolation

Plant Control
Optimization
DCV

0.2

0.0

Persistence
0.8
0.6
0.4
0.2
0.0
Total Electric
Savings
1e+06
2e+06
3e+06

Airflow Adjustment (Zone)

AHU Optimal Start

Thermostat Lighting
Control
Setting

retrofit since RCx, training for site personnel and BAS
management type (Table 4).
RCx Training. Training site engineering staff to operate and maintain RCx measures is critical to preserve
savings from installed measures. Our analysis showed
that staff who received training immediately following the RCx process were more knowledgeable about
RCx measures and made continuous efforts to preserve
energy savings. It is important to incorporate post RCx
building operation training as part of the RCx process.
Major Retrofits. When significant retrofit activity or
building additions followed RCx, persistence suffered.
After major retrofits, buildings were often zoned differently, had variations in space use or occupancy. All these
factors made the building operation deviate from the
post-RCx building management plan, causing a loss in
RCx savings. Outreach discussions could identify when
building retrofit, significant HVAC equipment retrofit,
or BAS hardware or software retrofit are occurring.
Internal BAS Management. We found that persistence
was more likely to decrease when the BAS was managed

https://www.ashrae.org/

ASHRAE Journal - December 2019

Table of Contents for the Digital Edition of ASHRAE Journal - December 2019