MMX - ENGINNERING SYSTEMS
Direct/Indirect Savings: US$ 1,000,000 (2% EPCM contract value)
Potential Savings: US$ 1,500,000 (3% EPCM contract value)
Avoidable Losses: US$ 500,000 (1% EPCM contract value)
Other: + Return on IT Investment
+ Project Progress Control
+ Design Quality
- Waste of Resources
+ Schedule Adherence
MMX was created in 2005 and quickly became an important mining company controlling several iron ore mines in Minas Gerais and Mato Grosso do Sul states (Brazil) and a particular port infrastructure.
In 2012, MMX launched its largest expansion program, the expansion of the Southeast System through the project known as Serra Azul, initially composed of a new mine and iron ore processing unit with capacity of 15MTA. The objective was to start production as fast as possible to take advantage of the high prices.
The challenging schedule for this expansion program forecasted the start-up of the processing unit in just 18 months, with a superposition between construction and engineering of 12 months. It was said to be a "FAST-TRACK" schedule where engineering and construction would follow almost in parallel with minimal separation of 2 to 3 months from draft to building. Instructions were not save resources to achieve the schedule. This determination was put to work and in just 40 days almost 200 professionals and top-notch facilities had been mobilized around a special task-force which acted independently within WORLEYPARSONS, the chosen EPCM contractor.
Seven months after, however, detailed engineering work seemed stuck, particularly in the execution of 3D model. Initially, WORLEYPARSONS involved some of it international specialists in order to understand what was going on, however without much success. In the end, IATEC Plant Solution was hired to diagnose and finally solve the problem.
One of the first problems detected was financial. Software histograms were flat, instead of typical hill shaped with a peak somewhere close the middle of project. As a result, an enormous amount of leased 3D seats were requested every month (23) and never used (15 on average), causing losses of -US$50,000/Month. Immediately, IATEC Plant Solutions recalculated all software histograms considering planned work and started to closely follow software demands which fell approximately to a half.
Expensive training, configuration and support services directed to third-party suppliers were eliminated or internalized resulting in immediate savings of US$ 250,000.
Overall software and hardware budget was cut to a half, to US$750,000 from US$ 1,500,000, what could represent higher savings if US$500,000 had not been already spent.
Once financial matters were equated it was time to tackle the technical part. Still gathering information about what was happening in the project, one thing attracted attention of IATEC Plant Solutions. While 3D review sessions showed at meetings were flawless, with all elements in their exact position, the same did not happen when it was cold extracted from PDS. This could only mean that elements within work database were in the wrong position and the 3D review session was post-processed to adjust positions.
Beyond the time wasted managing element positions every time the 3D review session was updated, there was an even worse problem that designers were working without minimal reliable references (or with uncertain ones) because models were not all centralized in the database. Another problem was that several 3D review sessions with different dates and configurations circulated across the project causing a great amount of uncertainty regarding what had been done or not. Not to mention the almost 400 empty models which made almost impossible to find the right references to work with or the backup problems created by such model dispersion across the project network.
Quickly, IATEC Plant Solutions brought all elements within the project database, eliminate "alternative" model locations in the network, and using a proprietary technology unified 3D review sessions in a unique source which was automatically updated every 12-hours with no manual intervention, leveling the information level across all project and avoiding the circulation of outdated information.
This measure also made every designer to work directly into the database (and no longer in "pirate models") and always access the 3D review session to check for references, preventing them from eschewing the hard but necessary interdisciplinary job.
Next, another activity of critical importance to the project organization was to create a practical, reliable and evidence-based model progress measurement process. Departing from a unique methodology created by IATEC Plant Solutions, a new model progress measurement process based on database information was created and quickly implemented uncovering the discrepancies between actual progress and former measurements based on a quite subjective approach.
One of the key advantages of this new methodology was that for the first time the design level of each model element could be inspected through the design review session anytime.
Once those basic system management changes were deployed it was time to tackle the productivity of the design team and bring the 3D modeling back on track with project schedule.
A constant complain from designer to explain the low productivity was the low performance of the workstations, which took a long time to open the models (sometimes hours) and frequently crashed during design work. Surprisingly, however, project had purchased the best workstations available on the market at the time, the powerful DELL T7500 on their top versions (US$15000/each), in clear contrast with other projects managed by IATEC which currently used standard desktops (US$ 2000/each) with absolutely no performance complains.
Inspecting the project database, IATEC PLANT Solutions uncovered several problems. Among them operational system configuration mistakes, license mapping problems, oversized reference files (up to 180Mb), empty and corrupted models and time-wasting modeling procedures which over-detailed models (i.e.: bolts, nuts, nameplates, fins, tire grooving...) which not only made modeling work to sluggish (x10 average time) but also represented a useless drag to workstation performance.
Summing up, it was invested more than US$250,000 in hardware while US$50,000 would be necessary, an irreversible loss of at least -US$200,000 to the project. Also, it was consumed at least 5000MH modeling hours to produce the equivalent to only 500MH, spending approximately US$150,000 in a task that should not had cost more than US$15,000, inflicting an additional loss of -US$100,000 to the project.
Another common mistake in the project was the usage of TEKLA steel-frame references that were too heavy and without intelligence. IATEC Plant Solutions implemented a procedure to import those models in a intelligent way where only information was transferred (instead of the heavy graphics) thus making the database lighter and more reliable. All this without causing any trouble to the steelwork discipline which as a result become more even integrated to the other project disciplines.
Specification and coding problems created great uncertainty and heavy re-work for piping, Electrical and Instrumentation and were also targeted by IATEC Plant Solutions.
Codification system was paired with corporative SmartPlant Materials (SPM and SmartPlant Reference Data (SPRD). It was also configured in both systems automatic translation for long and short material description to enable procurement and shop worked seamlessly under the same database.
Doubts regarding material configuration were clarified with potential suppliers bringing to a close months of debates among project stakeholders (i.e.: clients, engineering, procurement and assembly) what resulted into a more confident project progress.
IATEC Plant Solutions also detected discrepancies among the many engineering systems specifications used in the project which frequently opened room for constant complaints and payment retentions.
For example, isometrics format initially required in the project specifications made it difficult to use automatic generation because of unnecessary field typically used for manual isometric drafting. Adapting the system instead of the formats would cost a substantial amount of money and time and make isometrics less automatic, and thus, less reliable.
IATEC Plant Solutions proposed and negotiated a new format with the client, one that was widely used in the construction which could be made fully automatic.
Originally, it was suggested to use an expensive and complex 3D pipe support system (Pelican Forge), which were estimated to cost US$50,000 to configure and implement, not counting on the licenses (US$2500/month/each). Instead, IATEC Plant Solutions used a well tested procedure to modeling and naming pipe supports which allowed to extract the most out of the resources available on the 3D system at no cost.
IATEC Plant Solutions late participation on the project not only helped to recover the schedule but also resulted in direct and indirect savings in the order of +US$ 1,000,000 (2% Contract value). In case IATEC Plant Solutions had been mobilized right in the beginning of contracts additional savings of +US$ 500,000 could have been achieved.
Due to financial problems, Group EBX, the controller of MMX, decided to paralyze the project in May 2013, while it was at 50% project detailing progress and 3D design right back on schedule.