Srinivas Katipamula, Ph.D. - cacx. · PDF fileAnalyzing Interval Data Using the Energy...
Transcript of Srinivas Katipamula, Ph.D. - cacx. · PDF fileAnalyzing Interval Data Using the Energy...
Analyzing Interval Data Using the Energy Charting and Metrics (ECAM) Tool
1PNWD-SA-9511
Adapted from a presentation by:
Srinivas Katipamula, Ph.D.Staff Scientist, Battelle Pacific Northwest Division
Bill Koran, P.E.Senior Engineer, Quantum Energy Services and Technologies
Presentation Outline
• ECAM Capabilities
• Analysis of Interval Meter Data with ECAM
2PNWD-SA-9511
• A tool to assist with analysis of building performance data
• Facilitate the types of analyses that energy engineers, commissioning providers, and building operators might need to make
• “Semi-automation,” not complete automation of the analyses
What is ECAM?
3PNWD-SA-9511
• Because analyses is not completely automated, the tool is flexible, supporting additional analysis without limiting any normal spreadsheet functions
ECAM Version 2.0 New Features
• Excel 2007/2010 compatibility
• Full time history chart, filterable to selected months or days
• Load profile as box plots
• Load duration chart (histograms/frequency distributions)
• Chart to check input schedule
• Matrix charts
4PNWD-SA-9511
• Matrix charts
• Metrics by daytype, occupancy, and month
• Data summaries
• Additional scheduling periods for startup and shutdown
• Integrated user guide/help
• PNNL re-tuning charts (See http://buildingefficiency.labworks.org/large.stm for more details)
ECAM Detection/Diagnostic Features
• Detection
– Tool does not provide detection directly
– Assists with performance tracking
– Provides the ability to investigate the relationships between variables
5PNWD-SA-9511
between variables
• Diagnostics
– Tool does not provide diagnostics
– Future versions may provide some limited diagnostic capabilities, but this is not the focus of the tool
– General intent is to maximize the user’s ability to benefit from whatever data is available
Five Easy Steps
1. Select data from existing spreadsheet
2. Map points
3. Create schedules
6PNWD-SA-9511
3. Create schedules (optional)
4. Input energy project dates (optional)
5. Create metrics and charts
User’s Data — Example
DateTime
Outside
Dry Bulb
325
COMMON
CHWS
326
COMMON
CHWR
328
CHILLER
1 AMPS
342 COM
CDW
SUPPLY
343 COM
CDW
RETURN
345 A
CHILLER
ON
414
SCHWP-5
KW
415
SCHWP-6
KW
416
SCHWP-7
KW
8/23/04 12:00 AM 67.4 41.9 48.5 226.1 69.2 77.3 100 0 5 0.8
8/23/04 12:15 AM 67.5 42 48.7 224.3 69 77.1 100 0 4.8 1.3
8/23/04 12:30 AM 67.3 41.6 47.6 206.9 68.9 76.3 100 0 4.7 1.6
8/23/04 12:45 AM 67.1 41.8 47.7 206.7 68.5 75.8 100 0 4.6 1.3
8/23/04 1:00 AM 67.1 42.2 47.6 195.9 68.2 75 100 0 4.7 1.4
8/23/04 1:15 AM 67 41.8 47 193.8 68.1 74.8 100 0 4.6 0
8/23/04 1:30 AM 67 42 46.2 168.1 67.6 72.9 100 0 4.7 0
7PNWD-SA-9511
8/23/04 1:45 AM 66.8 41.9 45.3 152.1 67.5 71.9 100 0 4.6 0.6
8/23/04 2:00 AM 67 42.2 45.3 150.2 67.8 71.9 100 0 4.6 0.6
8/23/04 2:15 AM 67.1 42 48.5 0 67.6 74.6 100 0 4.8 0.9
8/23/04 2:30 AM 67.1 41.6 47.7 0 68.8 75.7 100 0 4.5 0.5
8/23/04 2:45 AM 67 42 48.4 0 68.2 75.1 100 0 4.7 0.8
8/23/04 3:00 AM 67.1 41.6 47.5 0 68.4 75.1 100 0 4.5 0.9
8/23/04 3:15 AM 67.2 41.7 47.5 0 68.3 74.8 100 0 4.7 0.8
8/23/04 3:30 AM 67.4 42.4 47.9 0 68.2 74.4 100 0 4.6 0.8
8/23/04 3:45 AM 67.4 43.5 48.6 0 68.3 73.8 100 0 4.7 0.9
8/23/04 4:00 AM 67.5 45 49.4 0 68.1 73 100 0 4.8 0
Step 1: Select data
1a. 1b.
8PNWD-SA-9511
1c. 1d.
Step 2: Map points
9PNWD-SA-9511
User’s points are mapped to common point names as perUser’s points are mapped to common point names as per
A Specifications Guide for Performance Monitoring SystemsA Specifications Guide for Performance Monitoring Systems
Step 3: Create Schedule (optional)
10PNWD-SA-9511
Forms are similar to eQuestForms are similar to eQuest
Step 4: Input Energy Project Dates (optional)
11PNWD-SA-9511
Step 5: Create Metrics or Charts
12PNWD-SA-9511
Preprocessing of Data
• Time stamp disaggregation
• Daytyping
• Occupancy
13PNWD-SA-9511
• Pre/Post Energy Project
• Normalizations
• Calculations
ECAM Metrics Creation Capabilities
• Can be normalized:
– Building area (e.g. W/sqft)
– Cooling tons (e.g. kW/ton)
– cfm (e.g. Watts/cfm)
• Can be filtered by:
– Year
– Month
– Pre/post time periods
In ECAM, metrics are typically data In ECAM, metrics are typically data averaged over a particular time periodaveraged over a particular time period
14PNWD-SA-9511
– cfm (e.g. Watts/cfm)
– gpm (e.g. gpm/ton)– Pre/post time periods
– Daytype
– Time of day
– Occupancy
– Weather conditions
– Status of equipment
– CombinationsAll filters and many normalizations All filters and many normalizations are automatically set up by ECAMare automatically set up by ECAM
Automatic Creation Of Additional Fields For Metrics
• Equipment Status
• Demand (kW)
• Chilled water tons
• from demand (kW) or
current (amps)
• from current (amps)
approximate calculation
• whenever flows and
temperatures are available
15PNWD-SA-9511
• Watts per square foot
• CFM per square foot
• kW per ton
• gpm per ton
temperatures are available
• for all electrical demand
points (kW)
• for all air flow points
• for all related points
• for all related points
Interval Data Visualization
• Load profiles
– Raw data
– Averages
– Pre/post
– By daytype
– Calendars
• Scatter charts
– Raw data
– Averages
– Pre/post
– By daytype
– By occupancy
16
– Calendars – By occupancy
– By time of day
– Choice of independent variable
– Multiple points
Filtering capabilities are providedfor all types of charts
Load Profile by Daytype
2.5
3.0
3.5
4.0
4.5
Avg
Ele
cM
tr_W
att
s_p
erS
F
Weekday
Year (All)
Month (All)
MonthYr Sep 2006
Weekday (All)
17PNWD-SA-9511
0.0
0.5
1.0
1.5
2.0
12:0
0 A
M
2:0
0 A
M
4:0
0 A
M
6:0
0 A
M
8:0
0 A
M
10:0
0 A
M
12:0
0 P
M
2:0
0 P
M
4:0
0 P
M
6:0
0 P
M
8:0
0 P
M
10:0
0 P
M
Avg
Ele
cM
tr_W
att
s_p
erS
F
Saturday
Sunday
Holiday
Weekday (All)
Day (All)
Holiday (All)
5degBin (All)
1degBin (All)
TempRng (All)
Filtering Options
Calendar Load Profile
October
Monday Tuesday Wednesday Thursday Friday Saturday Sunday
1 2
3 4 5 6 7 8 9
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
6
8
6
8
6
8
6
8
6
8
6
8
6
8
18PNWD-SA-9511
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31
0
2
4
0
2
4
0
2
4
0
2
4
0
2
4
0
2
4
0
2
4
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
3-D Load Profile
7/1/20067/2/2006
7/3/20067/4/2006
7/5/20067/6/2006
7/7/20067/8/2006
7/9/20067/10/2006
7/11/20067/12/2006
7/13/20067/14/2006
7/15/20067/16/2006
7/17/2006
7/18/2006
7/19/2006
7/20/2006
7/21/20064.00
5.00
12:00 AM
19PNWD-SA-9511
12:00
AM3:00
AM6:00
AM9:00
AM12:00
PM3:00
PM6:00
PM9:00
PM
7/21/2006
7/22/2006
7/23/2006
7/24/2006
7/25/2006
7/26/2006
7/27/2006
7/28/2006
7/29/2006
7/30/2006
7/31/2006
0.00
1.00
2.00
3.0012:00 AM
2:00 AM
4:00 AM
6:00 AM
8:00 AM
10:00 AM
12:00 PM
2:00 PM
4:00 PM
6:00 PM
8:00 PM
10:00 PM
7/1
/20
06
7/2
/20
06
7/3
/20
06
7/4
/20
06
7/5
/20
06
7/6
/20
06
7/7
/20
06
7/8
/20
06
7/9
/20
06
7/1
0/2
00
6
7/1
1/2
00
6
7/1
2/2
00
6
7/1
3/2
00
6
7/1
4/2
00
6
7/1
5/2
00
6
7/1
6/2
00
6
7/1
7/2
00
6
7/1
8/2
00
6
7/1
9/2
00
6
7/2
0/2
00
6
7/2
1/2
00
6
7/2
2/2
00
6
7/2
3/2
00
6
7/2
4/2
00
6
7/2
5/2
00
6
7/2
6/2
00
6
7/2
7/2
00
6
7/2
8/2
00
6
7/2
9/2
00
6
7/3
0/2
00
6
7/3
1/2
00
6
4.69-5.00
4.38-4.69
4.06-4.38
3.75-4.06
3.44-3.75
3.13-3.44
2.81-3.13
2.50-2.81
2.19-2.50
1.88-2.19
1.56-1.88
1.25-1.56
0.94-1.25
0.63-0.94
0.31-0.63
0.00-0.31Surface Chart
Box Plot Load Profile
20PNWD-SA-9511
3.0
4.0
5.0
6.0
Avg
Ele
cM
tr_W
att
s_p
erS
F
During Changes
Before 11-15-06
After 1-24-07
Scatter Chart by Date Range
21PNWD-SA-9511
0.0
1.0
2.0
0 20 40 60 80 100 120
AHU_TempOa
Avg
Ele
cM
tr_W
att
s_p
erS
F
Excel won’t create
Scatter Charts based on
PivotTables (PivotCharts).
ECAM includes
Scatter Charts based on
PivotTables.
Scatter Chart by Occupancywith Binned Weather
800
1,000
1,200
Av
g E
lecM
tr_k
W
22PNWD-SA-9511
0
200
400
600
20 30 40 50 60 70 80 90 100
1degBin
Av
g E
lecM
tr_k
W
Occ
Unocc
INTERVAL METER DATA ANALYSIS EXAMPLES
23
ANALYSIS EXAMPLES
Example: Load Profile by Daytype
24 Hour Occupied
24PNWD-SA-9511
24 Hour Occupied Schedule
Example: Load Profile by Daytype
Morning
Building occupied
(6 AM)
Unoccupied/ shutdown
•
25PNWD-SA-9511
Morning startup
Example: Typical Office Building
50% load reduction
between peak and nighttime
Long shutdown
period
26PNWD-SA-9511
5 day occupied schedule and unoccupied on
weekends
Example: Good Office Building
80% Peak Reduction
27PNWD-SA-9511
Good startup/shutdown
(6 AM/6 PM)
Building not occupied on weekend
Example: Comparing data before and after improvements
50% reduction
15% Peak load reduction
28PNWD-SA-9511
Additional 30% reduction at night
Comparing Consumption from Two Identical Buildings
29PNWD-SA-9511
Comparing Regression Analysis from Two Identical Buildings
30PNWD-SA-9511