By Bharat (Bob) Shah
Part 1 of this article, which appeared in the April issue of Hydrocarbon Processing, discussed the inception of piping and instrumentation diagrams (P&IDs), their development and maturity, and the challenges in between. The causes, scheduling and limitations of changes to P&IDs, and the Management of Change (MoC) process will be discussed here.
CAUSES OF LATE CHANGES
It is difficult to categorize or group the many reasons for routine or late changes to a P&ID. Some causes are listed here.
An unreasonable schedule. Contractor management will commit to unrealistic schedules to win a project. Sometimes, schedules are based on awarding the contract by a certain date, and then the award is delayed but the dated schedules remain unchanged. The engineers on the ground are stuck with that shorter schedule.
One of the most important items on the entire project schedule is the completion of the P&ID. The work of all engineering disciplines and their group’s deliverable schedules and staffing plan are based on this date. If P&IDs are delayed, job-hours are burned; if the delay is significant, personnel brought to the project must be de-staffed and the entire project schedule slips. Process engineering will need to provide justification for late changes.
P&ID development takes time. More people can be put to work on them, but this introduces the law of diminishing returns.
Poor scheduling. Schedules for an engineering, procurement and construction (EPC) project will contain thousands of items. P&ID schedules are watched more closely than anything else because the rest of the project work depends on them. One problem is that for large projects, several hundred P&IDs will have conflicting schedules showing start dates, internal reviews, client reviews, hazard and operability (HAZOP) studies, approved for design (AFD) and approved for construction (AFC) issues, etc. All of these P&IDs cannot have the same schedule; they must be divided into several groups of units or systems and each must have separate and staggered schedules.
Identical schedules with different contractors. When a large project is divided into two or more teams executed by different contractors in different parts of the world, it is inevitable that the teams will end up with different completion dates. When one team finishes its work, the work of the other team may still be evolving. Late developments by one team can have a big impact on the other team, resulting in P&ID changes.
Too few engineers. Engineers can get overloaded handling too many systems and too many P&IDs. P&ID reviews with clients, HAZOP reviews, etc., occupy a large number of job hours. Without careful planning, adequate staff to carry out non-HAZOP engineering becomes an issue. Typically, staffing should allow for four P&IDs per day for client review and four P&IDs per day for HAZOP review.
Inadequate or inexperienced supervision. Without adequate supervision, work quality suffers. Systems, such as utilities and offsites, do not get the deserved attention and suffer the most. In fact, because these are the open-art technologies, everyone comes up with their own ideas.
If the supervisor does not have the experience of having worked before on such systems, design will suffer from deficiencies. Cases have been noted where neither the working engineer nor the supervisor had the correct experience to execute the work. Omissions and errors were found—later rather than sooner—and fixing them forced others to redo their work and, of course, led to schedule delay.
Inadequate definition of work. If the work scope is poorly defined, changes will continue to happen over an extended period of time. A well-defined and detailed process design basis and scope of work can reduce this risk. Once again, utilities and offsite units, which are open-art technologies, suffer from such needed attention. In some cases, the design basis is copied from a type of facility that may not directly apply to other facilities. An experienced engineer will recognize this and try to convince the owner to change it.
Project philosophies. In an ideal scenario, project philosophies are established at the beginning of front-end engineering design (FEED); in reality, project philosophies continue to evolve during FEED and often change during the EPC phase. Once again, various parties are involved in writing these philosophies and gaps, conflicts and omissions can persist among the philosophies and/or between them and other project documents. Like with the design basis, a philosophy may also be a direct copy from a facility that is completely different than the current project and therefore may not be completely suitable. Additionally, a new party from the owner’s side may be assigned to the project and rewrite a philosophy. Various project philosophies that affect equipment and P&ID work are listed here:
- Relief, flaring, venting and blowdown
- Process control
- Operation and maintenance
- Design margin
- Material selection and corrosion management
- Emergency shutdown
- Process isolation
- Line sizing
- Metering philosophy
- Commissioning, startup and handover
- Overpressure protection
- Drainage philosophy
- Safety philosophy
- Fire and explosion.
Gaps and conflicts. No matter the effort, gaps and conflicts will exist in various project documents. Project-defining and scoping documents include an environmental impact statement (environmental assessment statement), project design basis, process design basis, FEED or EPC scope of work, FEED or EPC contracts, and engineering scope of work, among others. For FEED, there may be preliminary PFDs; during EPC, there will be FEED PFDs and P&IDs that typically are the basis of project cost of +/– 10% accuracy.
It is important to remember that multiple parties have produced these documents and may not have consulted with each other or seen each other’s documents. As end of FEED approaches, significant changes will be implemented in some FEED drawings; however, not all FEED drawings can be updated in time. If such last-minute changes are not separately documented, conflicts will show up during EPC.
Additionally, industry standards, practices and codes must be met. Timely identification and resolution of gaps and conflicts can minimize changes late in the game.
Interface failures. Most projects involve multiple parties. A billion-dollar project can have as many as 25 third-parties spread around the world. Some projects have extremely integrated interface between parties. A clear definition of philosophical and dimensional interfaces is of absolute importance (e.g., who is doing the engineering and design, purchasing and responsible for construction). This can be easily delineated in P&IDs.
Such big projects have dedicated and experienced location specific interface managers to keep all parties informed of progress and the needs affecting others. Not all parties will have impact on P&IDs, but surprises happen. A scope of work change or a design development by any third party can affect several other third parties.
Late inputs from outside parties. Outside parties are not present when P&IDs are being developed. For example, plant operation teams may get inundated with project documents for review and comments. They may be knowledgeable, experienced and have significant comments to contribute, but they may not realize the urgency of gathering and communicating those comments. In some cases, comments never make it to the contractor. At the same time, third-party operations and maintenance are busy with routine work so they may not pay timely attention.
Just as a contractor employs numerous specialists, owners also have their own subject matter experts (SMEs). While contractor specialists may be assigned full time to the project—or at least available on call—the same may not be true for the owner’s SMEs. They may be spread too thin to provide expertise to multiple projects, so their input may come too late. A contractor cannot do much other than document that they have sought the input from owner’s specialist in a timely manner.
Late information can also originate from lessons learned from previous projects or existing facilities. Lessons learned rarely come in a batch; they tend to trickle in regularly during FEED and EPC phases, arriving verbally in meetings, by email, conference notes, an old PowerPoint presentation, in the form of a pdf., spreadsheet, Word file, database, a consultant’s report, etc.
Keeping track of these numerous documents can be frustrating. A new owner’s representatives may arrive at a project at a late stage and bring their own new set of lessons learned. Licensors continue to gain experience from their similar units operating elsewhere and will insist that proposed changes be implemented. This results in late P&ID changes, which can be costly to implement and must be carefully considered because the proposed changes may not apply to the ongoing project.
Communication gaps. Examples of communication gaps that occur all-too-frequently include:
- Messages are not received
- Paper copy or emails are not forwarded to the affected parties
- Key personnel are not invited to meetings/consultations. Action items from meetings and reviews are not followed up with and may arise again
- HAZOP action items may be unclear.
FEED phase vs. EPC phase. In many cases, the FEED phase is carried out by one contractor and the EPC phase by another. When a challenging issue arises in FEED, the contractor may say this should be sorted out in the EPC phase; and when an issue comes up in EPC, the EPC contractor may claim that it should have been settled in FEED or it is not within the EPC scope of work. Regardless, the issue must be addressed.
HAZOP. HAZOP forms a significant and important part of a project. Regulators, project financers, clients and contractor policies require a complete HAZOP. Normally, a FEED HAZOP will identify action items that must be resolved in the EPC phase. Regardless, an EPC HAZOP is still required.
Scheduling a HAZOP study is vital, particularly because so many non-resident parties are required to attend. If a project contains a licensed unit, a licensor must be present for the HAZOP study, and complex vendor packages require a vendor package HAZOP that must be scheduled after the vendor drawings have been issued and adequately reviewed. If a third party is employed to facilitate a HAZOP study, it is preferred that all HAZOP studies are completed at one time.
In addition to open HAZOP action items, a separate open items list is normally maintained by the owners, with the help from the FEED contractor, and should be addressed in the EPC phase. This list may contain challenging items and last-minute design developments/changes that cannot be incorporated in various FEED documents. An estimate of cost may be made at the end of FEED but must be verified by the EPC contractor.
While the project is between the FEED and EPC phases, the owner’s team will be working on many open and new issues. As such, there will be scope change at the beginning of the EPC phase in all projects. Sometimes, the changes require significant development time and affect costs and schedule.
Splitting the crowded drawing. As a P&ID is developed, it becomes crowded. In its early phases, there are few instrument bubbles on the P&ID. By the time it is fully developed, it is completely covered with additional piping, equipment and instrument bubbles. The flowsheet draftsperson has difficulty fitting all the additions cleanly, so the P&ID can become crowded and unreadable.
Often, resistance exists to splitting the drawing because the affected P&ID is referenced in numerous other P&IDs, dozens of other project drawings and possibly those of other contractors, all of which must be changed. Weeks later, the P&ID is even more crowded and the only option is to split it. Split drawings may not get the adjacent sequence number, and operations might have to deal with these illogical sequence numbers.
Operating manual. As the EPC project approaches its completion, writing the operating manual is one of the last activities carried out by process engineers. This time-consuming activity must be properly planned. As the process engineer writes the manual to start, operate and maintain the plant, the design may require adaptation and, therefore, P&ID change, which is evaluated for priority criteria. In many cases, the home office effort is shut down and any engineering and design changes are handled by the field engineering team.
Management of change (MoC) process. Any change past the approved for design (AFD) or issued for design (IFD) requires approval/agreement from all parties involved. The sign off process is known as MoC. Each project may develop its own format for MoC, but the intent is the same. A one-page form will have a log number, describe the change, the reason for the change, cost/schedule impact, etc., as well as an approval signature block for all parties. A marked P&ID with the proposed change may be attached to the form, as well as a sketch and any supporting correspondence. All parties confer weekly to review, accept or reject the proposed changes. It is not uncommon for a particular proposed change to remain unresolved for weeks. The approved change may be highlighted with a cloud on the master P&ID. All approved change documents are collected in a binder that is maintained next to the master P&ID set for reference.
Project management is always tempted to freeze the P&IDs early, even when full development is incomplete. Therefore, development must occur with MoC in place, and therefore, too many MoCs are generated and P&IDs are covered with dozens of MoC clouds, some overlapping. Sometimes, an MoC must be retracted later as more data becomes available. Often, people get tired of these weekly MoC approval meetings, and key people begin to send engineers and designers in their place who have less experience and/or are unfamiliar with the issues. This can be embarrassing when a third party carries out a project audit and wants to know the P&ID status.
Regardless, good reasons remain for P&ID changes through the MoC process. A common question is what kind of P&ID changes can be made after the P&IDs are issued for design. Reasons for a justified late change to a P&ID are listed below.
In many cases, a HAZOP study will be required for approved MoC, which can be logistically difficult.
- Personnel safety— This is rare because the design has gone through numerous reviews by this time. However, such a change should not be ruled out.
- Equipment safety— This is also rare, but possible. Sometimes the impact of information received from the vendor causes change outside the equipment. A detailed discussion with the equipment vendor before the purchase order is placed can avoid such change.
- The current design will not work— The party proposing the change must make a case for why the current design will not work. Usually, this is caught early but can happen due to the aforementioned reasons.
- Non-compliance with the scope/specification—This can be caused by errors, omissions or late interpretation of the scope/specification.
- Non-compliance with regulations—This can be common, as regulations can be complex and sometimes conflicting. Local regulations can be more stringent than the central or federal regulations. Errors, omissions and misinterpretations can also be factors.
- Violation of a specific industry standard—An industry standard may have been overlooked or misinterpreted due to inadequate supervision of an inexperienced engineer’s work or a simple oversight.
- Change previously requested/agreed but was not implemented. This is mainly due to communications failure (e.g., a change requested weeks ago did not filter down to the proper party in time).
- Changes resulting from unresolved HAZOP action items— HAZOP comments must be resolved and incorporated in P&IDs before they can be issued as AFD or IFD. Often, not all HAZOP comments will be closed in time, but the P&IDs must be issued AFD. Affected parts of P&IDs are shown with clouds indicating a hold on the design for that part. No other discipline works on that part until the hold is removed. Therefore, there cannot be a large number of holds or clouds on the P&IDs in AFD/IFD issue.
Understanding the potential magnitude of the hold items is necessary. Management must take the initiative and ensure that the involved parties resolve the holds. Ignoring these issues can result in late changes to designs. HAZOP reviews can generate hundreds of comments requiring action, many of which are benign or have minimum impact on the ongoing design. For example, alarms may be added, a single transmitter can be changed to two out of three transmitters, or independent level transmitters may be added as a result of HAZOP comment. These are generally easy to implement and have minimum cost or schedule impact. Some comments will require action that is included in operating manuals, while others may require the design contractor to verify that a certain feature is included in the vendor package. Occasionally, an additional process simulation may be called for a specific situation. Some comments will require adding a detail or clarification notes to P&IDs.
Plant operations are required to attend the HAZOP study. The HAZOP comment may require an additional feature or more automation to the system. Operations will obviously support this, while the project team may resist as they have to pay for such changes.
For a proper HAZOP, many project documents must be available, including related PFDs with heat and mass balances, other connecting P&IDs, material selection diagrams, equipment data sheets, control philosophy (narrative), utility summaries, facility plot plan, area/unit/system plot plan, piping specifications, etc. These documents are fully listed in the HAZOP charter, which should be an early deliverable for the project. Sometimes, the parties responsible for providing these documents have not finished them. Without such supporting documents, some HAZOP facilitators refuse to start the HAZOP, while some will try to proceed without all of the documents. The project engineer responsible for the area should ensure that the required documents are being developed and will be ready for HAZOP. In the absence of the required documents, HAZOP will be improperly executed, resulting in a large number of action items and numerous last-minute changes to the P&IDs.
While most HAZOP comments will be resolved, many will remain open for weeks or even months. On lump sum turnkey projects, one discussion involves who will pay for the proposed change. On large projects, weekly meetings are held to approve and resolve HAZOP comments until they are resolved and closed. A project schedule may have a line item and date for resolving all HAZOP items. This may prove unrealistic. Ideally, items such as HAZOP should be scheduled as closures of 90%, 95% and 100%. The last 5% will prove to be the most time-consuming.
Unforeseen impact of vendor data. The failure to secure sufficient vendor data at the time of equipment purchase can cause changes. Once the purchase order is placed, the vendor will not reply until the vendor data is due (usually 6 week–8 week); by then, it is too late in many cases for ongoing design. Secure all necessary information from the vendor prior to placing the purchase order, including preliminary information on package footprint; package isolation valves; utilities required with rates, conditions, frequency and durations; waste stream produced and contaminants therein, etc. Some vendor packages must be located a minimum or a maximum distance from another system or equipment that is not provided by the same vendor. Failure to obtain this information results in many unplanned changes. One of the largest contributors to this challenge is the lack of communication with vendors and suppliers before placing a purchase order. By the time the approved vendor data arrives, it is too late.
Specific request by the company/owner despite cost and schedule implications. The client is paying the money and can change what it wants. Sometimes people in the plant will initiate change that is too late for the project, so the decision lies between the client’s project manager, who has to pay for the change, and the plant manager, who has to operate the plant. In some cases, this issue can linger until the project manager forces the resolution.
Takeaways. For any process plant, the entire project design originates from P&IDs that show all equipment and hundreds of other elements needed to build, operate, control and maintain the plant and produce the intended products. Development of P&IDs is a long process with continuous input from multiple parties. Late changes to P&IDs can result in costly overruns and schedule delays, so the reasons for proposed late changes must be understood and changes must be controlled.