W13_OAN_Pareto Priority Index

  1. Problem Definition

In this week Blog Posting, Author will analyze alternative method of tank repair. As we now, Storage tank is our critical facilities.  As Oil and Gas price now struggling, company has cost reduction campaign. Define the alternatives for project cost reduction and specify the priority of project that will be execute. Therefore, we must find the best alternative method to maintain reliability of our storage tank, especially old storage tank.

  1. Development of Feasible Alternatives

Feasible alternative project to maintain old storage tanks are:

Glass Fiber Reinforced Polymer (GFRP) Lining

Re-bottoming include cathodic protection

Plate doubling include cathodic protection

3.Possible Solution

Generate estimate cost, estimate saving and probability of success.

Estimate Saving based on fuel throughput per day, compare to construction for new tank (12 months project duration) and operation live of new tank (20 years).

  1. Glass Fiber Reinforced Polymer (GFRP) Lining (90 days project Duration – 5 years life time)
  • Total Cost                             : Rp. 400 Million
    Tank Cleaning
    Sand Blasting
    Fiber Reinforced
    Lining
  • Estimate Saving : Rp. Rp. 27 Billion
    Project duration
    Operation live

2. Re-bottoming include cathodic protection (180 days project duration – 5 years life time)

  • Total Cost : Rp. 1 Billion
    Tank Cleaning
    Repair all bottom plate
    Sand Blasting
    Cathodic Protection
  • Estimate Saving : Rp. 18 Billion
    Project duration
    Operation live

3. Plate Doubling include cathodic protection (135 days project duration – 2 years life time)

  • Total Cost : Rp. 300 Million
    Tank Cleaning
    Bottom plate assessment
    Plate doubling
    Sand Blasting
    Cathodic Protection
  • Estimate Saving : 16 Billion
    Project duration
    Operation live

Calculate Pareto Priority Index (PPI), Equation that will be use is as follow

Table 1: PPI Calculation Result

  1. Selection Criteria

Cost reducing project that have biggest PPI will be selected

  1. Analysis and Comparison of the Alternatives

Glass Fiber Reinforced Polymer (GFRP) Lining project has the biggest PPI value among other, according table 2 the project ranking as follow:

Table 2: Project Priority

  1. Selection and Preferred Alternatives

Based on above analysis, Glass Fiber Reinforced Polymer (GFRP) Lining project is selected due to has biggest PPI (216)

  1. Performance Monitoring and the Post Evaluation of Result

To select project more accurately, if it is available historical information can be use to estimate standard times and resources. Whenever update data is available, calculation should be re-run to validate selection.

Refrences

  1. Six sigma daily (2014). The Six Sigma Approach to Project Selection. Retrieved from: http://www.sixsigmadaily.com/implementation/the-six-sigma-approach-to-project-selection
  2. Selecting Projects where Six-Sigma can add value (2014). Retrieved from: http://unpan1.un.org/intradoc/groups/public/documents/arado/unpan020936.pdf
  3. W13_AL_Pareto Priority Index
    Retrieved from https://kristalaace2014.wordpress.com/2014/05/21/w13_al_pareto-priority-index/

 

 

W12_OAN_Car Selection using AHP

  1. Problem Definition

In this week blog posting, Author would like analyze choosing a car, using Analytic Hierarchy Process.

  1. Development of Feasible Alternatives

There are 2 alternatives car to choose, using 5 criteria which are Safety, Design, After Sales Service, Fuel Consumption, and Performance. Price will be set aside until the benefits of the alternatives are evaluated.

  1. Car A: Rp. 380 Milion
  2. Car B: Rp. 410 Milion
  3. Car C: Rp. 360 Milion
  4. Car D: Rp. 390 Milion

 

  1. Possible Solution

The information above is arranged in a hierarchical tree as below:

Figure 1: Hierarchical tree

  1. Selection Criteria

Table 1 will be the relative score:

Table 1: Relative Score

Car with the highest benefit – Cost Ratio will be selected.

  1. Analysis and Comparison of the Alternatives

Based on table 2, the relative score will be:

Table 2: Score – for alternative solution

Table 3: Average Value

Table 4: Weighted Matrix Comparison

Table 5: Consistency Vector

Table 6: CI Table

For Matrix 5 is 1.12.

We do same steps for each criteria.

Table 7: Rank based on Criteria

The alternative ranking is the product between alternative ranking matrix and criteria ranking.

Table 8: Alternative vs Criteria Matrix

  1. Selection and Preferred Alternatives

After alternative ranking has been defined, we have to compare between benefits and costs (price) using benefit – cost ratio.

Table 9: Cost Normalized

Table 10: Benefit – Cost Ratio

Based on Benefit – Cost Ratio, Car B has the highest ratio. Therefore, Car B will be Author choice.

  1. Performance Monitoring and the Post Evaluation of Result

In order to have more accurate result, pairwise comparison matrix can be done by surveying the experts or anyone who has good understanding of the problems.

Refrences

  1. Using AHP in Decision Making in Engineering Applications : Some Challenge
    Retrived from http://bit.csc.lsu.edu/trianta/Journal_PAPERS1/AHPapls1.pdf
  2. AHP Tutorial.
    Retrieved from https://pdfs.semanticscholar.org/7e27/b5a124c2e6829e1ff0d3e1279c2dbc9ebe2a.pdf
  3. W5_WW_Analytical Hierarchy Process
    Retrieved from https://garudaaace2015.wordpress.com/2015/03/28/w5_ww_-analytical-hierarchy-process-2/

 

 

W12_UDS_Emerald Blog Posting Learning Curve

  1. Problem Evaluation

We are already in week 13 AACE preparation course; this period is in middle of the course duration. From the beginning until now only blog posting task that have a real and constant target every week. When we make planning time duration (BCWS) 14 weeks ago, we just use our feeling without any formula or technique. This time we want to evaluate our blog posting learning curve to predict our next hour spends. It will be on the track or not (between early and late BCWS) also help us in rebase line our target.

Figure 1. Emerald AACE Blog Posting productivity

  1. Development of feasible alternatives

Refer to engineering economy book and Humphreys, they are two learning curve methods that we can adopt. Those two methods are:

  1. The Unit Linear Learning Curves (ULC)
  2. The Cumulative Average Linear Learning Curve (CUMAV)

In this small research, we analyzed our blog posting data in last 13 weeks to predict our next blog posting productivity, it also help us to get some picture of our learning curve in blog posting.

  1. Development the outcome for each alternative

This small research methodology is compare two learning curve formula ULC and CUMAV which more appropriate with our behavior blog posting productivity. Regression analysis will use as comparison rating index of both learning curve formulas ULC and CUMAV.

  1. Selection of criteria

There are three criteria of the best alternative in this method, there are:

  1. The learning rate (s) < 100%, the lower the better
  2. Coefficient Determination (R²) the higher the better
  3. Learning Curve Exponent (n) < 1, the lower the better.

The chosen alternative should have learning curve fits most of the criteria.

  1. Analysis and comparison of the alternative

Because we use group data (each week team productivity) in this research, first we should made lot unit to determine Lot Mid-Point (LMP). LMP formula that we are used is in the below :

Figure 2. LMP Equation

After LMP we should determine Average Unit Hour (AUH) of each Lot by dividing the total actual hours spending per lot to the lot size. The equation form default is exponential, it is will be easier and fit with our regression analysis method if we transform it Logarithm Equation below:

Ln (Yx) = Ln (T1) + b x Ln (X)

With definition:

  • Ln(Yx) = natural logarithm of each unit hours spend to produce X unit
  • Ln (T1) = natural logarithm of the first unit produced
  • Ln (X) = natural logarithm of unit X produced
  • b = the exponential value associated with the slope

For transform reason we use historical data of blog posting actual hour, the table is below:

Table 1. Blog Posting Historical Data

All data being calculated to following table for regression analysis pupose:

Table 2. Regression Analysis Calculation

There is no outliers’ data on table above, all data in between LCL and UCL line. In generate regression of each method (ULC and CUMAV) we use excel data analysis and get the summary below:

Table 3. ULC and CUMAV Comparison Result

After we get the equation of both method ULC and CUMAV, we can predict of the rest blog posting hour spending, using additional assumption team will produce 7 blog each week so prediction table result is below:

Table 4. Blog Posting Hour Projection

  1. Alternative selection

From the comparison result table above has shown that The Unit Linear Learning Curves (ULC) Method is considered to be the favorable method to apply, rather than Cumulative Average Linear Learning Curve (CUMAV) Method using Heuristic LMP (Lot Midpoint) since its R² has shown higher rate than the CUMAV. The ULC’s R² value of 0.186 explain that the equation for estimating purpose is best fits with the data being analyzed. While the value of s = 0.72 represented the learning rate of Emerald AACE Team in Blog Writing Project. This number also tells us that Emerald AACE’s skill improvement in doing Blog Writing Project from W1 to W13 have been increased by 28%. The n value, indicate that Emerald AACE Team underspending in this project by (-0.28) and (-0.02) worth.

  1. Performance monitoring & Post Evaluation Result

Using appropriate Learning curve method of our team (ULC) can give us a projection of our rest blog writing project especially when we want to rebase line our BACS. It is projection can guide us to plot our new BACS and maintain our CPI on the track. We can do the same evaluation for the other project in couple of weeks, when it already enough data, because the other project target not start in the beginning of preparation course.

References

  1. Sullivan, G. W., Wicks, M. E., & Koelling, C. P.(2014). Engineering economy 16th Edition. Chapter 3 – Cost Estimation Techniques., pp.110-113. Prentice Hall.
  2. Humphreys, G. (2014). Project Management Using Earned Value (Third Edition). Chapter 22 – Learning Curves., pp.435-441.Humphreys & Assoc.
  3. Planning Planet. (2017). Acquiring Man Power For The Project. Retrieved from http://www.planningplanet.com/guild/gpccar/acquiring-manpower-for-the-project
 

W14-ABM-Developing the BCWS recovery curve using IEAC


Problem Definition

Further to my Week 13 blog and the development of new baseline schedule, the development of a revised BCWS (early and late curve is required).

The new BCWS curve must assess the extent of balance works and the forecast cost required to complete. To support this process we will now look at the current Estimate at Completion figures and use these estimates to help establish recovery BCWS curves.

Calculating the IEAC can be performed using a number of different methods  and this blog will look at each method and its suitability in supporting the development of recovery BCWS figures/curves.

Feasible Alternatives

EAC data will be assessed at both the Programme and Project levels using the following methods;

  • IEAC1 = ACWP + ((BAC – BCWP) / CPI)
  • IEAC2 = ACWP + ((BAC – BCWP) / SPI)
  • IEAC3 = ACWP + ((BAC – BCWP) / CPI * SPI)
  • IEAC4 = ACWP + ((BAC – BCWP) / ((0.2 * SPI) + (0.8 * CPI))

Development of the Alternatives

Table 01 below is taken from the NDIA/GPC and outlines the assumptions and considerations when using each of the 4 forecasting tools.

Table 01

Each of the techniques will be assessed using project data from Week 13 as per table below. i.e An EAC will be assessed for each individual project / deliverable.

Table 01 – Project Performance Data (week 13)

Selection Criteria

The following criteria will be used to assess the preferred method of establishing the EAC and using to establish a revised BCWS curve for the recovery schedule

  • EAC takes account of Scope which was previously “underestimated” for  and is corrected for balance activities
  • EAC is realistic but also reduces the risk of reporting further increases i.e. pessimistic.
  • Forecast costs accurately reflect the remaining scope

Comparison of the Alternatives

EAC results for each method and level are provided under table 2.0

Table 02 – IEAC Results

The original BAC for all 6 projects was estimated at $20,700 however IEAC calculated using the 4 methods above predict a cost overrun in the range of +20,060 to $62,936 above BAC.

Analysis of these results is as follows

  1. Project 6 – Bid Project:Project 6 is yet to commence and appears to be influenced by a imbalance in ACWP & BCWP figures taken from Week 0 milestone split. CPI and SPI figures for this project are not considered to be reflective of the actual remaining works and it may be more suitable to ignore this section from the EAC calculation and continue to rely on the original BAC.
  2. This imbalance from week 0 split is also observed in projects 2 and 5 where hours recorded within week 0 account for the bulk of current ACWP despite relatively low BCWP recorded. This skews the CPI and SPI figures
  3. IEAC3 – A result of $83,636 is not considered realistic and appears to be influenced by the late starts made on cheat sheet and Bid project (See above).
  4. It can be seen that there is a major difference between the EAC figures calculated at the programme level and those which have been calculated using performance data from each individual project.

Selection of Alternative

From the above, IEAC3 will not be considered as it is not considered realistic and appears to be overly pessimistic.

IEAC 2 is not considered appropriate given the impact of skew to SPI figures due to heavy reliance on SPI and limitations within later part of project.

Whilst IEAC 1 and 4 have produced values which are almost identical in value, IEAC 4 is considered the most appropriate given its partial consideration of SPI ensuring that figures are not overly optimistic.

However adjustment of the Total EAC to account for Project 6 (BID) will be made to take account of unrealistic EAC figures generated when using the reported CPI and SPI figures. This adjustment would be relevant to all options regardless. EAC for project 6 will be based on 100% of BAC.

Finally, to ensure that the accuracy of the EAC is maintained in all areas of the programme, the IEAC will be based on the cumulative EAC value for all 6 projects and not based on the programme level CPI and SPI figures. ($35,186)

With respect to the development of revised BCWS curve, the EAC values to be used are as follows;

Table 03 – IEAC FINAL

Performance Monitoring

Whilst adjustment of BCWS figures is not considered a regular event, the  monitoring of IEAC figures is a continual process and must be undertaken regularly. IEAC4 is recommended for use within weekly reporting against the new BCWS recovery curve.

References

  1. W09_SJP_Forecasts retrieved 5 November 2107 from https://js-pag-cert-2017.com/w09_sjp_forecasts
  2. Chapter  9.5 – Performance Monitoring Progress – Guild of project controls compendium and reference (CaR) | Project Controls – planning, scheduling, cost management and forensic analysis (Planning Planet).  Retrieved from http://www.planningplanet.com
  3. National Defense Industrial Association. (2014). A Guide to managing programs using predictive measures.