(or How I Learned to Stop Worrying and Love Performance Specs)
This document is an attempt to describe how today's accepted practice of construction document and bid RFP preparation takes place, and where there is an opportunity to move faster and reduce soft costs.
This is intended to cover the most common scopes in the most common buildings, which is the vast majority of the work that occurs. There will be some building types, measures, and/or other situations that don't fit perfectly in this framework.
Let's look at a steam or hydronic boiler replacement project as an example:
Owner hires engineering firm to prepare design and construction documents.
Engineer puts together design based around a particular boiler and burner combination.
This is typically called the Basis of Design, and it effectively sets a performance specification for what the project must include.
Effectively, the equipment schedule is the majority of their contribution to the project.
Not to mention, in many cases the Basis of Design is provided by a manufacturer's rep and just cut and pasted onto the drawing sheet.
The rest of the construction documents are often boilerplate, including generic text such as "the construction must comply with code," "follow manufacturer's installation instructions," and "drawings are schematic only; contractor is responsible for providing a full working system, including any necessary pieces that aren't shown." That's a fun one.
The project goes out to bid and someone wins the job.
We then enter the period of submittals and shop drawings, which is where the real detailed design work happens. Contractors submit the actual products they are planning to install on the project. These may be two things:
They match the Basis of Design.
Don't get excited and think this means the engineer's design was so specific and good that there were no alternatives. It usually means that particular contractor is already familiar with that brand of equipment and prefers working with it already. It's a scotch marine steam boiler or a Laars boiler plant—no one is winning prizes for originality here. Most of these scopes are so highly regulated, and owners and contractors are so risk averse (rightly so), that there's little variation to be had.
In any case, most of the installation details are still on the contractor.
They submit an alternative product. The engineer needs to review that submittal and provide comments back to the contractor. They can either:
Review the entire design with this new product in mind and update drawings accordingly. This rarely happens. It creates more work for them, which isn't in their budget.
Review the high-level performance data to make sure it meets the project goals. Provide comments back to the contractor to the effect of: "Product meets performance requirements, but coordination with the site dimensions and rest of scope wasn't reviewed. Contractor is responsible for ensuring this product will fit in the space and meet all other project requirements as necessary."
(Also the engineer most likely didn't think through all the installation details in the first place, so really only can review for equipment performance.)
At this point the owner may ask if there is any reason not to go with what the contractor submitted vs. the original Basis of Design. It usually comes down to what's going to be more reliable to operate and/or service, and that's an area where the contractors have greater insight than the engineers. Or what the contractor can get better pricing on. And that probably informed the products they submitted already, so likely to win out.
Submittal is approved and construction begins.
Construction is completed and inspected.
At the end of the project, the engineer usually has to revise the original drawing set to match what was actually built and file a post-approval amendment (PAA) with the DOB to match the final field conditions.
There is a need for a third-party engineer, or some other knowledgeable agent. The emphasis here should be more on third-party, and less on engineer. More like a referee.
The need is really for an inspector acting on behalf of the owner to gauge construction progress, to approve payment requisitions accordingly, and to make sure the work performs as expected and meets the owner's requirements.
You would think that all of this would fall into an engineer's normal role. But no! Over the years, whether due to cost pressures, abilities, or some combination of the two, most engineers have limited their services to providing the most basic installation and punch list inspections they can get away with.
Testing the equipment to make sure it performs as expected has fallen to a new role: the commissioning agent. This is really a huge backfire to the pressures of cost cutting over the years. Engineering costs were pushed down so far that the level of service decreased and the market came up with an entirely new role to fill that gap. Now owners have to pay for both engineers and commissioning agents for most scopes.
There is still a need to set owner requirements and performance specifications for the projects. Those are tasks generally associated with owner's reps and commissioning agents, and we still need them.
But all that time the engineer spent selecting equipment and drafting up voluminous plans with tiny print that few people read in any detail is time and money wasted. And the need for a post-approval amendment at the end of construction, in order to revise the original drawings to match what was actually installed, is also time and money wasted. That money can be saved without a meaningful change to the way the market works. We can call it "right-sizing soft costs" if we need to.
So much of the installation requirements for projects flow from code and the installation instructions. Engineers rarely, if ever, read the installation manuals and don't always know what's required or sometimes even how to apply the equipment properly.
Contractors are on site for the inspections (whether it's the manufacturer's rep for startup and warranty purposes, the DOB for boiler inspections, or the utility for gas piping inspections) and know first hand what is required to obtain the necessary approvals by the relevant authorities.
Engineers are rarely on site during those inspections. Often the utility won't even be responsive until a contractor is on board with the project and in the conversation. Engineers hold no clout in this part of the process.
If something doesn't work with the equipment after the installation, the contractor gets the first call. They want to avoid these callbacks and the associated costs, and so work to minimize risk in the installation.
All of this combines to a tight feedback loop for contractors; they are able to feed learnings from one project into the next very quickly.
Engineers experience a much slower feedback loop. A lot of the information they get is secondhand. And sometimes they get no feedback at all. The contractor may adjust things to pass inspection and/or resolve an owner complaint, and that information never gets back to the engineer.
When they do get feedback from projects—say, if the total cost comes in higher than their initial estimate—it takes a long time for that learning to show up in their other projects. All their previous reports and estimates were delivered in static PDFs, and there's no business reason for them to go back and update older reports.
One of the things we are building with Momentum is a better way to close the loop faster on project outcomes. Data about actual project costs and schedules will be captured and fed back into the tool so we can provide more accurate estimates and expectations to users.
Strategy Notebook
The Myth of Third-party Design
(or How I Learned to Stop Worrying and Love Performance Specs)
This document is an attempt to describe how today's accepted practice of construction document and bid RFP preparation takes place, and where there is an opportunity to move faster and reduce soft costs.
This is intended to cover the most common scopes in the most common buildings, which is the vast majority of the work that occurs. There will be some building types, measures, and/or other situations that don't fit perfectly in this framework.
The Boiler Replacement Example #
Let's look at a steam or hydronic boiler replacement project as an example:
During Construction and Post-construction #
There is a need for a third-party engineer, or some other knowledgeable agent. The emphasis here should be more on third-party, and less on engineer. More like a referee.
The need is really for an inspector acting on behalf of the owner to gauge construction progress, to approve payment requisitions accordingly, and to make sure the work performs as expected and meets the owner's requirements.
You would think that all of this would fall into an engineer's normal role. But no! Over the years, whether due to cost pressures, abilities, or some combination of the two, most engineers have limited their services to providing the most basic installation and punch list inspections they can get away with.
Testing the equipment to make sure it performs as expected has fallen to a new role: the commissioning agent. This is really a huge backfire to the pressures of cost cutting over the years. Engineering costs were pushed down so far that the level of service decreased and the market came up with an entirely new role to fill that gap. Now owners have to pay for both engineers and commissioning agents for most scopes.
There is still a need to set owner requirements and performance specifications for the projects. Those are tasks generally associated with owner's reps and commissioning agents, and we still need them.
But all that time the engineer spent selecting equipment and drafting up voluminous plans with tiny print that few people read in any detail is time and money wasted. And the need for a post-approval amendment at the end of construction, in order to revise the original drawings to match what was actually installed, is also time and money wasted. That money can be saved without a meaningful change to the way the market works. We can call it "right-sizing soft costs" if we need to.
On Feedback Loops #
So much of the installation requirements for projects flow from code and the installation instructions. Engineers rarely, if ever, read the installation manuals and don't always know what's required or sometimes even how to apply the equipment properly.
Contractors are on site for the inspections (whether it's the manufacturer's rep for startup and warranty purposes, the DOB for boiler inspections, or the utility for gas piping inspections) and know first hand what is required to obtain the necessary approvals by the relevant authorities.
Engineers are rarely on site during those inspections. Often the utility won't even be responsive until a contractor is on board with the project and in the conversation. Engineers hold no clout in this part of the process.
If something doesn't work with the equipment after the installation, the contractor gets the first call. They want to avoid these callbacks and the associated costs, and so work to minimize risk in the installation.
All of this combines to a tight feedback loop for contractors; they are able to feed learnings from one project into the next very quickly.
Engineers experience a much slower feedback loop. A lot of the information they get is secondhand. And sometimes they get no feedback at all. The contractor may adjust things to pass inspection and/or resolve an owner complaint, and that information never gets back to the engineer.
When they do get feedback from projects—say, if the total cost comes in higher than their initial estimate—it takes a long time for that learning to show up in their other projects. All their previous reports and estimates were delivered in static PDFs, and there's no business reason for them to go back and update older reports.
One of the things we are building with Momentum is a better way to close the loop faster on project outcomes. Data about actual project costs and schedules will be captured and fed back into the tool so we can provide more accurate estimates and expectations to users.