Reference no: EM131931970
Semester Design Project - Phase 2 ALL-LineX Superior Energy Performance
BACKGROUND
ALL-LineX produces "bucket" extendable arms for the telecommunications, electrical, and fiber optic businesses and are used by lineman that work on telephone/electrical poles. These units are very customized with some that are specifically designed for use in high voltage environments and require proper shielding and grounding precautions.
The plant manager, Sally Rider, has been tasked with implementing a corporate strategy to reduce their energy intensity by 10%/year for each of the next three years. Sally is quite skeptical of this goal and is frankly not all that interested in this effort since it's not part of her evaluation and bonus structure being comprised of costs, person-hours, quality, and delivery performance. She has asked you as the plant/industrial engineer to take responsibility for the effort.
Figure 1 gives an overall layout of the plant. There is a rather significant office area that is utilized 5-days/week from 7 am - 6 pm. There are occasional times when someone is working during the weekends or late at night and the lights and HVAC are adjusted to accommodate them. There is small warehouse area for raw materials and some replacement/repaired parts. This area is generally available at all times that the plant is in operation. There is a mechanical area where compressed air and other building facilities are located. The main area of the plant is operated 5-days/week for two 8-hour shifts. This is primarily some metal fabrication, testing, and assembly operations.
There is a paint line that consists of a WIP staging area, part loading/unloading staging, a 6- stage wash and part prep operation, a drying oven to dry parts, a manually operated paint powder spray booth, and a final cure oven. The paint line operates primarily 5-days per week, one shift per day, but at times they may be asked to extend their operation to a second shift or to the weekends during peak times.
Finally, there is a test location where scientists test and perform R&D on high voltage exposure to the bucket equipment (5-day work week 9-5). The high-voltage experiments operate sporadically over the 5-day week since this may impact electrical usage.
Because this is a union environment, the entire 300,000 sq. ft. facility is environmentally conditioned using 23 rooftop HVAC units.
![427_figure.png](https://secure.expertsmind.com/CMSImages/427_figure.png)
Superior Energy Performance:
As a young plant engineer, you've heard about companies that have participated in the Department of Energy's (DOE) Superior Energy Performance (SEP) program, and the corresponding ISO 50001 standard, that have recorded phenomenal improvement in energy performance. Measurement and verification of energy improvements is conducted by measuring "energy intensity." In general terms, "energy intensity is a measure of the energy efficiency of a nation's economy. It is calculated as units of energy per unit of GDP. High-energy intensities indicate a high price or cost of converting energy into GDP. Low energy intensity indicates a lower price or cost of converting energy into GDP." From an individual plant, energy intensity is a measure of the relative energy needed to produce a unit of goods sold.
Scope of Phase 2:
Phase 2 is primarily about constructing adequate multiple linear regression models for both electric usage and natural gas usage given the contextualized data from Phase 1 - including either elimination or approximation of up to four electric or gas data points based on feedback from Sally Ride's agent once the Phase 1 reports have been reviewed. The dataset should not include the last 6 months of data.
1. Using the "Correlation Analysis in JMP" document on eCampus, construct a correlation and covariance analysis of the following variables:
a. Electricity Usage
b. Natural Gas Usage
c. Sales (3-month Moving Average transformed data)
d. HDD
e. CDD
f. Total OSHA hours including Regular, OT, and Double OT
g. Total Regular-time OSHA hours
h. Total OT and Double OT hours
i. Total OT hours
j. Total Double OT hours
k. Repeat the correlation analysis using (a) - (e) but instead of items (f) - (j) use just the Paint line hours for items (f) - (j)
l. Repeat the correlation analysis using (a) - (e) but instead of items (f) - (j) use everything but the Paint line hours for items (f) - (j)
m. Repeat the correlation analysis using (a) - (e) but instead of items (f) - (j) use transformed variables for Paint line OT hours (x2) and Paintline Double OT hours (x2).
Using these correlation analyses, make recommendations about correlation and covariance of response and predictor variables for consideration of inclusion in the multiple linear regression analysis. Discuss the covariance of predictor variables and determine which variables will most likely result in problems with multi-collinearity.
2. Based on the results from part 1, conduct multiple linear regression analyses (see "Multiple Linear Regression for JMP" in eCampus) for both electric usage and natural gas usage using the predictor variables that would seem to make the most sense to include in the regression analyses - justify your inclusion of variables. Where it might make sense consider interaction effects as well as any squared variables. Run these regression analyses multiple times with different orders of effects and explain why these models might be different based on the correlation of the estimates.
3. Based on the results from parts 1 & 2 (including suitability for any interaction and squared effects), conduct two multiple linear regression analyses for both electric usage and natural gas usage using both a forward and backward stepwise regression analysis. Contrast the results from these two approaches and advocate for overall regression models for predicting electric and natural gas usages.
4. Conduct scatter plots of the residuals for both electrical and gas usage for the models advocated from part 3 and comment about the assumptions of constant variance and normality.
5. Construct 95% prediction intervals for new observations of the models from part 3 and then examine whether the data excluded from part 1 (last six months of data) fall within these intervals. Report on prediction errors of these 6 months and the suitability of the models for predicting future electric and natural gas usage going forward.
6. Using MMBTU as a common energy unit, determine the overall energy savings by using the Forecast Normalization method as described in the Superior Energy Performance M&V Protocol document.
Attachment:- Assignment Files.rar