In a Greentech Media featured article, Lucid’s Michael Murray untangles the messy and expensive issues around systems integration.
Making systems talk with one another is much more than just a hassle. It’s so difficult that it supports an entire industry called ‘systems integration’.
I recently asked 65 experts in facility management, energy efficiency and green building, “How much does system integration cost our industry as a whole?”
It’s a deceptively simple question.
When I use the term “system integration,” I am referring to the labor or hardware costs required to make HVAC systems, submeters, lighting controls, and other devices in buildings talk nicely with one another. To the uninitiated, systems integration seems like it ought to be a cake walk: If I can install a printer driver on my PC without a lot of hassle, how hard can it be to, say, pair Modbus devices over a local network? Or connect BACnet controllers with one another? Hasn’t technology solved that problem?
But for those of us who work on integration problems every day, making systems talk with one another is much more than just a hassle. It’s so difficult that it supports an entire industry called “systems integration.”
Many building operators spend unexpectedly large quantities of time and resources on integration, which usually means consulting fees to systems integrators or the big automation companies. If paying large sums of money to give systems integrators job security doesn’t give you a warm, fuzzy feeling, then you’re probably equally frustrated by what that money could be doing for facilities if it was put to work in more productive areas, such as energy efficiency.
Let’s start with a tangible example of the integration hurdle. An owner of several 200,000-square-foot buildings wants to install an electric meter in each one and tie it to the building automation system (BAS). This customer wants to track usage in greater detail than the monthly bill can provide. Eventually, this customer wants to explore reducing HVAC loads at certain peak times, but right now he simply needs to gather data. Sounds easy enough, right?
With a quick Google search, you can find an electric meter for $250 plus shipping. Let’s say labor is $500 for the electrician. It’s uncertain how much the BAS guy is going to charge, but a ballpark estimate would put it in the $250 to $500 range per building. How hard can this be? He just needs to run some wire, hook things up and push some buttons, right?
If you estimated the total cost of this hypothetical installation to be merely $1,000 or $1,250 per building, then you would be wrong. The BAS guy’s quote is astronomical. First, you need to add an “input board” to the field panel to accommodate the additional meter input. There is labor required to configure this properly. Then you need to “trend” the point in the BAS — this means actually storing the kilowatt-hour data on disk.
The BAS is not set up for trending. How to fix that? You’ll need to buy another computer to run the database. Okay, computers are $500 from Best Buy. Not a deal-killer. But now you need a Microsoft SQL server license, an expensive BAS software license, and, most expensive of all, engineering labor ($175 per hour for several days) to set it all up.
“Where did I go wrong?” you ask yourself. You decide to get a price quote from another BAS guy. It’s the same amount (or more). Frustrated, you conclude that getting submetering done is just expensive.
Sadly, the building automation industry was never designed to make integration easy. In fact, it was designed to do just the opposite.
The companies that manufacture BAS controllers have always relied on local dealers or distributors to market and sell their products. Local dealers won’t carry a product unless they can make money selling it. Hardware markups in the 40 percent to 50 percent range are nice, but they’re nothing compared to a situation in which every sale obligates the customer to purchase twenty to 40 hours of engineering time, which you can bill out at $175 per hour.
Not surprisingly, “easy-to-install” BASs can scarcely be found on the market. Whether you believe in planned obsolescence or think that’s just how business is done, your submetering project is foiled. Customers do not have access to “single-click” systems integration because the BAS manufacturers don’t want to damage their revenue lifelines (i.e., their local dealers). As one of my 65 experts aptly wrote, “BACnet is revolutionary as it relates to the normal standard of care for the industry.” This was the best backhanded compliment I’d heard in a long time.
Back to my survey. There were no published reports I found detailing the costs of systems integration for, say, submetering projects; sadly, even robust datasets such as Lawrence Berkeley National Laboratory’s ESCO database do not itemize cost data to that level. But nonetheless, my colleagues pointed me to these interesting statistics:
- An LBNL study of energy information systems found that BAS integration could cost $30,000, $40,000 or even $50,000 per system.
- A large corporation (with over 100 buildings) cited systems integration as one of its greatest challenges across its portfolio. With experience and diligence, the company successfully whittled down the integration time for a new BAS to approximately 24 man-hours. If only medium-size customers could all achieve similar results.
- Another LBNL study of the commissioning industry found that 69 percent of all commissioning costs are due to “investigation and planning,” a significant amount of which is data collection. If we conservatively estimate that 10 percent of commissioning “investigation and planning” costs are related to data connectivity with BASs, that means we spend $14 million per year unnecessarily because BASs are not easy to work with. And that is on commissioning alone, which is a tiny fraction of all BAS market activity in new construction and retrofit markets. (I’m assuming a U.S. commissioning market size of $200 million/year.)
Unfortunately, these statistics don’t fully answer my original question. Nevertheless, if you doubt that systems integration is a large burden on the industry, consider the number of companies that are financially supported by filling this niche. The Control System Integrators Association (CSIA), for example, has 453 member companies, and one of the largest multi-protocol integration platforms, Tridium’s Niagara (owned by Honeywell), has hundreds of dealers across all 50 states.
It probably doesn’t need to be said that more energy efficiency could be realized if our budgets weren’t being artificially siphoned by disappointing BAS technologies.