Johnson & Johnson, makers of hundreds of brand-name products from Band-Aid to Tylenol, operates 11 research laboratories worldwide under its Pharmaceutical Research & Development LLC (J&JPRD) subsidiary. Three years ago, the J&JPRD site in La Jolla, CA, was outgrowing its space and needed to double its square footage to accommodate about 600 scientists and staff. In the proposed expansion, the existing L-shaped building would be matched by another, and the two connected, yielding a total of 308,000 square feet surrounding a rectangular courtyard.

At the time, the original building was buying power under an energy services contract poised to expire. The facility’s managers needed an integrated solution for both the old and forthcoming new sections, naturally wanting what would prove most economical. Still fresh in everyone’s minds was California’s energy market fiasco, which had driven electric rates to a five-fold increase, just months before. As J&JPRD’s Mark Loukides recalls, in those days, “No one was really interested in doing long-term electrical contracts anymore,” and this spurred innovative thinking. Too, the notion of becoming grid-independent looked extremely attractive, as did the opportunity of having highly efficient combined cooling, heating, and power (CHP). J&JPRD’s thinking was, “Once we get a CHP system in, the [escalating utility rates] will no longer be relevant,” recalls Loukides, head of operations for the division. Also appealing was the environmental benefit.

However, CHP comes along for many as uncharted territory. Backup power generation is well known to many site managers and engineers, but gensets used for prime baseloading running 24/7 are another story. The idea of capturing nearly all exhaust heat and using it to provide hot-water service and run chillers—inherently appealing to most engineers as a concept—usually means piecing together a custom-designed system, and this can turn out to be something of a challenge and an adventure.

Electrical and Heating Load Assessment
First question: Precisely how much power should be installed? And what generation type is preferable? To address these and other issues, a detailed energy profile is required. In this instance the survey would encompass both the historical load data in the current building and projections on usage in the new. As Loukides recounts, average daily loads and yearly totals were thus calculated based on refined, hour-by-hour measurements. Such detail is indispensable, as it will be matched against utility-peaking rates in order to tally the full cost-avoiding value of cogen power.

In this particular seaside locale during warmer months, electrical loads increase beginning at midday, when power peak rates in California also climb dramatically. Thus, engine exhaust heat can be almost fully utilized by an absorption chiller. Obtaining good matches like these turns out to be the key to making projects go.

As for the heat load calculations, the site’s current senior engineer Bill Grantham, and others, relied primarily on measuring boiler therms. A bit surprisingly, though, despite La Jolla’s mild California climate, the coastal zone stays a bit on the chilly side, he says, “and we have more days when we heat the building than we cool it.” Once a comfortable temperature is reached each morning, though, further heating isn’t needed.

Taking this profile and load data, the study team considered various solutions. The most “enticing,” so to speak, at least at first, were fuel cells and microturbines, reflective of sexier and presumably more advanced technologies. However, in examining several actual field applications that were explored nearby, the reported results were mixed. The primary problem in many marginal CHP jobs often stems from a mismatch between the generating technology and the actual heat load that will be needed. Right-sizing on this point turns out to be more critical to the outcome than does the nature of the generator technology per se.

“Looking at four or five CHP applications,” Loukides recalls, “we found they were clearly using all the electricity, but in a lot of cases they weren’t using the heat [adequately]—and yet this assumption was built into the overall analysis. It’s like, if you’re wasting the heat, you’re wasting a lot of your energy profile.”

During this exploratory phase the team also met a well-seasoned onsite power developer, PID Engineering, conveniently nearby in San Diego. J&JPRD thus hired PID for expertise both on doing the load profile and to advise on hardware options. PID’s Tom Lamar remembers, “We gave them all the inputs they needed as far as annual fuel consumption, electricity generation, how much they’d still need to purchase from the utility, and rough cost of installation,” all with an open mind regarding possible solutions. Based on past experience though, the gut feeling was that reciprocating engine generator sets “would be the best way to go, as opposed to gas turbines, because of the heat load,” says Lamar, who ultimately became the project’s mechanical engineer.

J&JPRD’s spreadsheets eventually reached the same conclusion. Loukides adds, “We learned from others’ mistakes.” One was “what you don’t want to do is to fall in love with any technology ... and then go and try to force-fit your energy profile to make a fuel cell work [where it really doesn’t].” Again, he stresses, it’s critical to “understand your energy profile, then go out and evaluate your different technologies to see which one best fits your organization.”

Another valuable lesson to emerge from application site visits, says Loukides, is to “get a good maintenance agreement.” One implementation customer had bumped into an untimely surprise regarding the need to have engine overhauls every fourth year. This interval is perfectly normal for any engine churning out power 24/7—but, when the overhaul comes due, it will of necessity quadruple the baseline maintenance budget; this must be fully factored into life-cycle costing. The adopter was thus “kind of taken by surprise,” or at least hadn’t adequately considered recurring maintenance costs, Loukides remarks.

So, when J&JPRD sent out bid requests, it asked for full-cycle maintenance quotes. This would cover all levels ranging from routine weekly, monthly, and quarterly servicing up to major rebuilds, with everything pretty much pre-scheduled for that period. Also, by signing on for long-term integrated maintenance linked to guaranteed engine performance, the winning bidder—whoever that might eventually be—would have a big incentive to keep his engines in top running shape.

Engine Selection, Quick Payback
In the end, after weighing several technology packages the best choice turned out to be, indeed, the workhorse reciprocating engine. It offered the optimal combination of power density, efficiency, emissions controls, and usable heat, matching the profile.

The RE field was then narrowed to a few finalists in the 1.0-MW to 1.3-MW size range—that being the optimum for taking advantage of a $900,000 rebate available under the California Self-Generation Incentive Program (SGIP). Onsite power projects that qualify for this are generously underwritten, in order to help the state meet its burgeoning demand for electrical generation; in the formula, there’s also a heavy bias favoring “green.”

Two finalists remained. Tipping the decision in favor of the ultimate winner was its power density within a small footprint, and thus, J&JPRD selected two 1.1-MW, 16-cylinder engines from Cummins (i.e., 2.2 MW total).

PID drafted the integrated heating-cooling-power-train design, based on Cummins’ guidelines. Lamar recalls, “Cummins had packaged quite a bit within their enclosure. I didn’t end up having to specify a lot of equipment. There was already a cooling system available in the form of cooling towers that were supplying the existing building, and they were going to be providing for the new building [as well].” All in all, he says, the challenge “wasn’t a giant scope on the mechanical side.”

J&JPRD’s total installed cost for the engines and related elements came to “a few million bucks,” says Loukides. But not everything was fully itemized, because some HVAC aspects were rolled into the larger construction budget, and new elements were married up to existing circuits and hardware.

Again, though, a very big chunk of this sum was offset by the SGIP. Too, beyond this there was a six-figure cost avoidance achieved by eliminating backup generators that would normally be purchased or upgraded for a new building. The reason for not doing so is that if the two Cummins units happen to go offline, the local utility, San Diego Gas & Electric (SDG&E), can take over “seamlessly” (or nearly so), and thus effectively serve as backup standby power. And as an added reliability bonus, the onsite engines can power J&JPRD in island mode; even if SDG&E power fails, the lights and all systems stay on.

Both engines were initially budgeted to run (and in fact do run) almost full tilt, to provide around 90% to 95% of the facility’s electrical load—the 100% level being precluded under the interconnection tariff.

To obtain a permit, J&JPRD first had to demonstrate that the circuit wouldn’t back-feed power to the grid. Additional commissioning protocols were required under the SGIP.

Still another avoided cost was hot-water boilers and storage tanks; these were readily eliminated in the new building by CHP. The two engines’ paired exhaust-heat exchangers can take inflow circulation in the jackets and raise the temperature in a continuous stream; so, scratch a couple of boilers.

The outflow divides between domestic hot water, and depending on the season of the year, the balance travels either to a space-heating loop or to the new Thermax LiBr absorption chiller. The latter is neatly complemented by the continued service of the original electric chiller unit.

All in all, then, J&JPRD figures its avoided costs come to about $500,000.

As for savings on gas and electric billings, the company recently pegged this at about $1 million a year. Assuming a 5% annual increase in utility rates in future years, break-even should come, says Loukides, “after about five years.” Unfortunately, post-Katrina natural gas Btu costs are spiking (and were already climbing in 2004–2005 anyway, relative to electric rates); payback time frames are anyone’s guess.

Construction: “On Time, Under Budget”
Siting and housing the two generators optimally was comparatively easy on this job, as the whole site was undergoing expansion. A prime spot near a parking lot was cleared to house the engines; circuit connections run back to the building underground. The equipment layout on this expensive terrain was “tight but workable,” notes Lamar.

For noise abatement, an acoustic consultant helped design the enclosure, and the contract limits the allowable decibels. In fact, notes site engineer Grantham (who, of course, is closest to daily operations), the generators have proven surprisingly quiet. “More sound baffling was actually required on the cooling tower,” he says, and sound levels are completely satisfactory.

Doing the engine installation honors—along with inking that 10-year maintenance agreement—was the local Cummins dealership, Cummins Cal Pacific LLC (CCP), of El Cajon. Unit one thus came online in 2003 to handle the original building; unit two, powering the new one, was switched on early in 2004—a two-stage strategy that spread the SGIP qualification over two years to maximize the rebate.

Commissioning the absorption chiller (sometimes tricky) took a month to get online, and now, says Grantham, it works as smoothly as the electric chiller.

In sum, says Loukides, construction proceeded well, coming in “on time and a little under budget.” PID’s Lamar credits much of this to weekly meetings, a benefit of all participants being close by.

CCP—having a big stake in long-term system performance—is monitoring hundreds of parameters with onboard real-time sensors. It’s logging “in five second intervals,” reports CCP’s Brian Mayhall, general maintenance manager. PLCs and sensors keep track, he says, of “fluid pressures, temperatures, alternator bearing temps, SCR [selective catalytic reduction] system, flow rate of urea, temperature differentials coming in and going out, plant load,” and more—literally, “everything that can be monitored is monitored” (i.e., about 300 parameters). Datalogs are quickly accessed via a DSL line. If any indicator shows something out of whack, a pre-alarm condition is signaled so that corrections can be made. Sensors automatically page and dispatch a technician. This kind of heavily monitored array, Mayhall notes, is rapidly becoming standard in Cummins’ markets.

Glowing Performance
Output-wise, since their staggered commissioning dates, the two engines combined have yielded about 15 million kWh/yr of electricity; 367,844 therms of heat in the first operational year; and 1.6 ton-hr/yr of chilled water. All told, this equates to more than 90% of the facility’s total electric power and much of its heating and cooling needs.

Environmental benefits, too, have been impressive. The system’s cogeneration effect (i.e., recycling engine heat) reduces carbon dioxide emissions (a “greenhouse gas”) an estimated 3.2 million pounds per year, according to an EPA formula; that’s equivalent to 285 cars on local roadways. The electrical generation at 2.2 MW translates into 3.9 million kWh of electricity annually—enough for about 540 homes.

The latter effect means that SDG&E ratepayers also gain substantially from J&JPRD’s investment, by virtue of postponing the need for buying more electrical capacity—which explains the SGIP’s big subsidy.

How About Daily Operation?
“There is no operator,” says Loukides, meaning it literally. Thanks in part to all of the Cummins monitoring, the power plant “takes care of itself.” Grantham and others know how to shut it off if needed; but, apart from daily walk-arounds and eyeing a few gauges, he and his staff have little direct involvement with the engines.

Glitches associated with startup and commissioning are almost routine, and this case was no exception. For a time, spontaneous shutdowns were being triggered, and this required painstaking trouble-shooting. The problem turned out to be in the integration of the engines with the existing cooling tower—which now also was supplying the engines’ coolant; the loop proved more unexpectedly problematic due to resulting changes in water flow. Dips in pressure caused engine overheating—leading first to derating, and then shutdowns. The eventual fix came from installing a booster pump.

Mayhall observes, “These are complex systems, and what might seem to be the problem isn’t always the root cause.” He notes that cogen integration with an existing HVAC system “can be a definite pitfall.” CHP designs become more customized and are subject to pre-existing quirks. Troubleshooting takes time with so many variables.

Air-Quality Reporting
One operational aspect that has required some attention at times is emissions control, but the sensitivity here is perhaps more political in nature than technical. The local Air Pollution Control District (APCD) requires extensive monitoring—which, fortunately, and as noted above, is built in. Engines are equipped with the standard selective catalytic reduction equipment and urea-adding devices. Depending on circumstances, as many as 10 parameters must be logged and reported. In this particular case, as Cal Pacific’s Mayhall explains, monitors keep tabs on exhaust temperatures and load factors on each unit, auto-adjusting the urea flow rate based on these two (urea being added to the combustion stream to reduce NOx).

So far, the engines have easily stayed in compliance, but interactions with the APCD staff are still time-consuming; so, to make life easier for everyone, the decision was made to install backup NOx monitoring. This will also increase precision and ensure continuous monitoring in any eventuality. The APCD, notes Grantham, “does not like to hear any ‘excuses.’”

Another strategy to keep regulators happy, adds Mayhall, is to acquire your own equipment and do monthly spot checks, which he’s doing, he says, “to find out if we truly are meeting the emissions that our customers are permitted to,” instead of waiting for a surprise regulatory visit. Another tip: Have a maintenance person annually inspect the catalysts’ condition, “and if they’re getting plugged, or whatever, we’ll replace them, in most cases.”

Coming Next: Fine-Tuning, Role-Modeling
The system is roughly meeting the original theoretical expectations with respect to cogen efficiency. However, as Loukides notes, now that it’s been in operation awhile, its heat output is proving to be a bit higher than was forecast—meaning that additional utilization opportunities remain. Currently, five heat exchangers are in place (two on each engine and a fifth on the cooling-tower loop). Even with these, there’s still heat to be drawn off, so a sixth exchanger will be added sometime around the New Year. When it comes online, Loukides anticipates it will enable the decommissioning of two remaining boilers, “and reduce some of the cost of reclaimed water.” Overall exhaust-heat utilization will then leap from the current 75% rate up to very nearly 100%.

Also ahead, Loukides reports that J&JPRD is exploring addition of a 200-kW solar-photovoltaic system for peaking power. Though it’s not a critical need in terms of energy supply, solar power does reflects the company’s larger clean-energy commitment.

Speaking of which, after 18 months on the job, in late summer 2005, the J&JPRD site received a coveted LEED (Leadership in Energy and Environmental Design) certification from the US Green Building Council, a coalition supporting environmentally friendly building design and operations. Resource conservation and environmentally conscious features were installed throughout the building and run the gamut from water-use devices to lighting fixtures; but, singled out particularly in the award was the cogeneration efficiency, which, as already noted, reduces carbon dioxide emissions by many tons.

Besides receiving the LEED award, it is also notable that the La Jolla site has turned into something of a role model for other potential cogen clients. One recent site visitor from the neighborhood, for instance, was impressed to learn that the system’s grid-parallel generators essentially eliminate the expense of buying standby diesel gensets; this factor in itself would make the value proposition of cogen very attractive.

Moreover, notes Loukides, other J&JPRD units are being assessed as cogen candidates. Plans are underway to install CHP at one company facility on the East Coast and another in Puerto Rico. Engineers from both sites have visited La Jolla to acquire some hard-won lessons and insights, and they’ve been calling back for more. Other implementations are sure to follow. Loukides sums up that onsite cogeneration “is something that Johnson & Johnson has been looking at adopting as a technology in a much broader sense, not necessarily at every location ... but is certainly looking at putting, literally, in sites throughout the world.”

DAVID ENGLE, a writer based in La Mesa, CA, specializes in construction-related topics.

DE - January/February 2006