Report

RTF Lighting Standard Protocol Review of Hours of Operations, Controls Savings, and Variation Hours • Context: SR relies on interview used for hours with default as backup. How do our proposed defaults look? • Review measured hours versus default • Caveats – All logged data by some method from studies meant to ascertain hours – Unknown protocols for annualization – Unknown site sampling procedures & duration – Know little about building/occupant charateristics – Litlle know about sample selection • It would take substantial time mine for those unknowns Hours of Operations Datasets Used • CPUC & Itron. (2010). Small Commercial Contract Group Direct Impact Evaluation Report CALMAC Study ID: CPU0019.01. Retrieved from http://www.calmac.org/abstract.asp?id=2739 • CPUC Database for Energy Efficient Resources (DEER). (2010). Summary of 2008 DEER Measure Energy Analysis Revisions Version 2008.2.05 – 09-11 Planning/Reporting Version, Comparison of 2005 and 2008 EFLH for Lighting. Retrieved from http://www.deeresources.com/deer0911planning/downloads/DEE R2008UPDATE-EnergyAnalysisMethodsChangeSummaryV9.pdf • New York State. (2012). New York State Technical Resource Manual. Retrieved from http://www.aging.ny.gov/livableny/ResourceManual/Index.cfm • Bonneville Power Administration. (2011). BPA C&I Lighting Calculator. Retrieved from http://www.bpa.gov/energy/n/projects/lighting Default from Calculator Total = 27 default options Reviewed = 14 Reviews were completed on datasets that had over 4 quality data points. Descriptive statistics were completed on all analyzed default inputs. Data was cleaned by eliminating statistical outliers. See workbook for full statistical analysis. (tab: Stats Review of Hours) College or University Descriptive statistics Yearly Hours Count 46 mean 2,719.26 sample variance sample standard deviation 73,756.73 271.58 minimum 2137 maximum 3241 range sum sum of squares deviation sum of squares (SSX) 1104 125,086.00 343,460,518.00 3,319,052.87 Hospital Descriptive statistics Yearly Hours count 27 mean 4,088.89 sample variance sample standard deviation 1,070,182.26 1,034.50 minimum 2497 maximum 5900 range sum sum of squares deviation sum of squares (SSX) 3403 110,400.00 479,238,072.00 27,824,738.67 Lodging Descriptive statistics Yearly Hours count 38 mean 3,605.42 sample variance sample standard deviation 3,862,898.14 1,965.43 minimum 755 maximum 7884 range sum 7129 137,006.00 sum of squares 636,891,548.00 deviation sum of squares (SSX) 142,927,231.26 Manufacturing Descriptive statistics Yearly Hours count 37 mean 2,943.73 sample variance sample standard deviation 166,890.59 408.52 minimum 2250 maximum 3916 range sum sum of squares deviation sum of squares (SSX) 1666 108,918.00 326,633,216.00 6,008,061.30 Office <20,000 sf Descriptive statistics Yearly Hours count 26 mean 2,746.58 sample variance sample standard deviation 407,228.49 638.14 minimum 1556 maximum 3957 range sum sum of squares deviation sum of squares (SSX) 2401 71,411.00 206,316,517.00 10,180,712.35 Office >100,000 sf Descriptive statistics Yearly Hours count 28 mean 2,906.96 sample variance sample standard deviation 430,193.81 655.89 minimum 1647 maximum 3860 range sum sum of squares deviation sum of squares (SSX) 2213 81,395.00 248,227,591.00 11,615,232.96 Other Health, Nursing, Medical Clinic Descriptive statistics Yearly Hours count 14 mean 3,621.93 sample variance sample standard deviation 13,247.61 115.10 minimum 3468 maximum 3814 range sum sum of squares deviation sum of squares (SSX) 346 50,707.00 183,829,351.00 172,218.93 Restaurant Descriptive statistics Yearly Hours count 37 mean 4,036.19 sample variance sample standard deviation 701,765.32 837.71 minimum 2724 maximum 5000 range sum sum of squares deviation sum of squares (SSX) 2276 149,339.00 628,024,009.00 25,263,551.68 Retail Boutique <5,000 sf Descriptive statistics Yearly Hours count 15 mean 3,138.67 sample variance sample standard deviation 309,082.67 555.95 minimum 2402 maximum 4055 range sum sum of squares deviation sum of squares (SSX) 1653 47,080.00 152,095,584.00 4,327,157.33 Retail Supermarket Descriptive statistics Yearly Hours count 23 mean 3,844.17 sample variance sample standard deviation 845,342.79 919.43 minimum 1979 maximum 4964 range sum sum of squares deviation sum of squares (SSX) 2985 88,416.00 358,484,022.00 18,597,541.30 Retail Big Box >50,000 sf One-Story Descriptive statistics Yearly Hours count 24 mean 3,408.33 sample variance sample standard deviation 677,648.67 823.19 minimum 2338 maximum 4800 range sum sum of squares deviation sum of squares (SSX) 2462 81,800.00 294,387,586.00 15,585,919.33 Retail Anchor Store >50,000 sf Multistory Descriptive statistics Yearly Hours count 18 mean 3,214.44 sample variance sample standard deviation 310,169.20 556.93 minimum 2559 maximum 4057 range sum sum of squares deviation sum of squares (SSX) 1498 57,860.00 191,260,632.00 5,272,876.44 School K-12 Descriptive statistics Yearly Hours count 12 mean 2,372.75 sample variance 2,739.66 sample standard deviation 52.34 minimum 2291 maximum 2452 range sum sum of squares deviation sum of squares (SSX) 161 28,473.00 67,589,447.00 30,136.25 Warehouse Descriptive statistics Yearly Hours count 33 mean 2,874.73 sample variance sample standard deviation 158,175.95 397.71 minimum 2074 maximum 3522 range sum sum of squares deviation sum of squares (SSX) 1448 94,866.00 277,775,508.00 5,061,630.55 Some Conclusions • Significant range for hours within some building use types highlights the importance of the interview part of SRM • Use of default carries more uncertainty in some use types than others • Default hours for some building use types should be revised • Test of simplest reliable needs to include how often default is used • What is correlation between logged hours and reported business hours? Hours of Operations Conclusion • Hours of Operations varies across building types and within building types. It depends on the building, occupants, type of work, and location within the building (e.g.: office vs. break room vs. computer room vs…..) • Because of large variation within building types it will be difficult to use defaults and estimate accurate operating hours • Even within the same utility, different programs report different operating hours for the same building type. • A regional primary study would most likely return the same uncertainty and not be worth the $$. Control savings as a % Issues: – Most analysis, case studies focus on full lighting retrofits not just control upgrades – There are large variations in control only reported data – Type of lights, control settings, and operating hours all effect results – Location of controls effects performance 3 studies are presented to show the variability LBNL Meta-Analysis from Lighting Controls • Focus on commercial buildings by building type • 240 savings estimates from 88 papers and case studies • Control types evaluated = Occupancy, Day lighting, Personal tuning, Institutional Tuning, and multiple types • Report website: http://efficiency.lbl.gov/drupal.files/ees/Lighting %20Controls%20in%20Commercial%20Buildings_ LBNL-5095-E.pdf LBNL Meta-Analysis Conclusions •Table 7 page 15 •These are the data points used in the calculator •These points are the average of all studies analyzed •These numbers include studies that are calculations and actual installation LBNL Meta-Analysis Conclusions Large Standard deviations indicated (+ - 20%) uncertainty and varying results Actual installations are not broken up by building type Rensselaer Polytechnic Institute review of Occupancy Sensors • Evaluation of Documented and undocumented studies by building location type • Evaluation identifies ranges and average savings • Most data evaluated reported prior to 2001 Rensselaer Polytechnic Institute review of Occupancy Sensors Findings M & V of Day lighting Photocontrols • 5 Buildings evaluated (office, school, manufacturing, medical, warehouse) • Looked at the direction of Office Spaces • All locations based in Idaho • Report website: http://www.idlboise.com/pdf/papers/201001 11_Final_Photo%20Controls.pdf M & V of Day lighting Photocontrols Findings •Results from monitoring period during regular operation hours •Direction of Offices has large significant impact on savings •Variations in site savings provides uncertainty •Numbers and savings % are different from the LBNL study Controls Savings Conclusions • Documented variations are large • Occupancy variations, directional variations, and space type variations, add to uncertainty • Small samples of installations do not accurately represent the population of installations • Most calculations overestimate actual savings