Legionella: Be Proactive, Not Reactive
Is Legionnaires' Disease (LD) still a viable threat? Definitely. On Feb. 5, 2004, operating engineers at a plant in Michigan were looking for the possible cause of an LD outbreak. At the same time, a hotel in Baltimore was being visited by LD. Less than a month later, an Oklahoma City hotel had more than 50 guests fall victim, most of them high school students.
Is Legionnaires' Disease (LD) still a viable threat? Definitely. On Feb. 5, 2004, operating engineers at a plant in Michigan were looking for the possible cause of an LD outbreak. At the same time, a hotel in Baltimore was being visited by LD. Less than a month later, an Oklahoma City hotel had more than 50 guests fall victim, most of them high school students. The list goes on.
Yet I've encountered professional engineers, who when the issue comes up, complain that LD is no longer an issue. Frankly, this is because news of all these cases doesn't reach them, and they assume it's a non-issue. But designers of engineered building systems and their clients—the owners—need to be concerned. In fact, there are only two options: Take the reactive approach, and deal with LD after the fact; or implement proactive strategies to reduce the risk of LD before it strikes.
Waiting to react
In a reactive mode, what does one do? After an outbreak, money cannot be an issue. Experts must be brought in to test everything, and all systems must be sanitized according to guidelines defined by the U.S. Centers for Disease Control and Prevention (CDC), ASHRAE, the Cooling Technology Institute or other recognized organizations—all at a significant cost.
Moreover, the inconvenience to staff and facility occupants can be horrendous. During and after this traumatic experience, facility staff must look at options to prevent recurrence. Some facilities take a long time to recover from the stigma of adverse publicity. The Bellevue Stafford Hotel in Philadelphia comes to mind. The hotel, which was hosting a Legionnaires' convention in 1976, was the site of the earliest documented LD case and thus gave the disease its name.
Going on the offensive
The good news is that one can be proactive. With such an approach, authorities weigh the consequences of an LD outbreak before it happens—and take action beforehand. This was, in fact, the case at the James Square Nursing Home in central New York. The administrative staff considered the following options for their two potable hot-water (PHW) loops:
Keep the circulating temperature above 140
Install a dual 5-micron filter system that continuously feeds chlorine to the system to maintain 0.5 parts per million (ppm) residual at 100%.
Install an off-the-shelf copper/silver (Cu/Ag) electrode device to feed copper and silver ions to the PHW loop.
Install a generic Cu/Ag injection system.
The first approach requires installation of thermal pinch valves at every faucet to prevent scalding. Besides cost, drawbacks include possible valve malfunctions and staff inconvenience.
The filter-system option is used by many health-care facilities. However, one needs staff and a chemical consultant capable of understanding oxidation-reduction-potential controllers and probes. Furthermore, such probes have a tendency to interfere with a frequency that depends on the source of potable water (PW). Sensitivity of probes also decreases over time if they are not properly maintained.
Cu/Ag electrodes are gaining popularity and bear the imprimatur of many researchers. Efficacy and cost, however, are issues. The installed average cost per unit is about $32,000. Furthermore, most of these systems require a spare electrode for the monthly cleaning process. Electrode replacement costs are in excess of $3,000. Another drawback is the lack of filtration and backflow preventers.
But an even stronger argument against this system is its performance history. In the late 1990s, Rohr, Senger, Selenka, Turley and Wilhelm conducted a four-year study of a German hospital employing the technology, base but discovered poor results. It should be noted that German codes do not allow silver to be in excess of 10 parts per billion (pob), which may explain their results.
But poor results also were encountered in the United States at University Hospital in Iowa City, Iowa, where an outbreak was experienced even though they, too, had a Cu/Ag ionization system.
This brings us to option four, which I call the "generic" approach. This alternative, which was actually applied at the James Square Nursing Home, involves adding the salts of copper and silver as pure solutions to the make-up water of the PHW loop. One installs a properly sized contractor water meter into the cold-water feed. This water meter will make a contact for every X gallons of water that flows through it. The water is then subjected to a bank of bag filters at 5.0 microns. The filters address dirt/debris in the line, removing a potential food source. It also captures conductor at 5.0 microns. After the filters, chemicals are injected at a rate commensurate with the actual make-up water following the system-load curve to achieve the focused objective of 400 pob copper and 40 pob silver. Since piping is also copper, problems with this parameter will arise; however, one need only "tweak" the pump to adjust the copper level—and only the copper that is being injected—without affecting the silver. Once the correct copper level is achieved, the chemical-feed pump will continue to keep it there, varying only as the level of system copper—due to pipe corrosion—varies. Monthly testing is recommended to insure that Cu/Ag levels are within the 400/40 pob range.
A refinement to the generic system could opt for a 5-micron filter system at the main and ultraviolet light applied to the PHW feed line, then to the Cu/Ag. This will remove many conductor and dirt, while addressing free-swimming Legionella bacteria (LB).
The electrode investigation continues
Not convinced? Let's revisit ionization-electrode technology and even more detailed research conducted on its effectiveness. A study in 1995 was undertaken to ascertain efficacy of ionization electrode technology over an 18-month period. The results were presented at the annual meeting of the Water Works Operators Assn. of Pennsylvania. Four nursing homes were surveyed. Two appeared to be free of LB; all four, however, tested positive for amoebae—indicators and amplifiers of LB. The worst-affected of these facilities opted for the installation of a Cu/Ag ionization device. Results were monitored over an 18-month period, which ran from January through June of the following year. For the record, the release rate of the silver and copper ions with this particular technology is regulated by a microprocessor.
The study found that in the nursing home with the Cu/Ag device, mean levels of Cu and Ag were inconsistent throughout. The target silver level was 10 to 30 pob—out of step with the accepted focus of 40 pob. The target copper level was 100 to 400 pob, likewise out of step with the accepted focus of 400 pob. The accepted focus should be constrained within reasonable plus or minus swings. This is reminiscent of a chemistry and biology "finagler" maxim: If the results do not meet the goal, change the goal to meet the results. In other words, the study proved what the laws of physics and chemistry said could not be accomplished.
It was also noted that during the course of the study, observers discovered debris-scale buildup on electrodes, which they suspected may have contributed to the odd results. From this point in the study on, a spare electrode was inventoried and switched monthly. An unusual spike was also recorded during the Oct.%%MDASSML%%Dec. period. The study's observers attributed this anomaly to the fact that the device was boosted to 3.5 amps from its original 2-amp operation to address the low-ion issue. However, an inexplicable spike was also experienced in the final month.
Despite these anomalies and problems, the study did find that the application of copper and silver had an immediate and dramatic effect on the number of sites that tested positive for LB. However, complete eradication was not achieved. During the course of the study, it was theorized that all LB capable of being killed, in fact, were destroyed. It was further hypothetized that what remained were cells, via Darwinian Laws, that adapted resistance to heavy metals.
In the end, the study lent credence to the conclusion that Cu/Ag ionization-electrode technology produces LB reduction, not eradication—if the parameters are not focused. Remember, the counts represent colony-forming units (cfu) per milliliter (ml); thus, one cfu equals one bacteria or bacterial clump per ml, or 1,000 per liter, or 3,540 per gallon. While some risk assessments may suggest a low risk, keep in mind there was still a presence of LB and amoebae.
Can one do more? Yes! For copper, the target is 100 to 400 pob. If the alloyed electrode was capable of meeting the focused objective, one would not see such wide fluctuations of ions being added. The original alloy mix when ionization devices first hit the marketplace was 90% copper to 10% silver. When this ratio failed to meet expectations, it was changed to 70/30 and may hover today at 60/40.
In the case of this nursing home study, if one removed the high spikes they'd discover the 10%%MDASSML%%30 pob criteria was not met, thus the Cu/Ag technology is not a panacea effective against all bacteria. Some companies make outrageous claims such as, "We know that silver kills some 650 bacteria, while the average antibiotic kills approximately half a dozen. Silver is the only effective way to kill certain viral strains such as Legionella pneumophilia, the virus responsible for Legionnaires' Disease."
No silver bullet
The nursing home study dispels some of these claims. In fact the figure below shows that for the heterotrophic bacteria that require nitrogen and carbon for metabolism, heterotrophic plate count (HPC) stayed positive throughout the test. In other words, one must conclude no effect by Cu/Ag on the capability of bacterial species growing on nutrient agar containing nitrogen and carbon.
Additionally, the Coliform reading was, as expected, low in chlorinated water. What was unexpected were the positives: the sites for amoebae were high, and since LB utilize amoebae as hosts for amplification, this isn't good.
As for chlorine content in the cold make-up and hot water (average 39.5
This study, though limited in scope, did show that Cu/Ag has apparent merit in Legionella control, but more importantly, it showed that the $32,000 ionization device failed to provide continuous residuals within the guidelines provided.
Another study, based on a written survey of 16 hospitals using Cu/Ag ionization devices, reported better news, noting that after installing the ionization units, nosocomial LD did not reappear. At the same time, the results of 253 hospitals polled by CDC in the anonymous National Nosocomial Infections Surveillance (NNIS) survey offered some contrasting results. In this survery, it was revealed that supplemental potable water-decontamination systems were installed in only 20% of hospitals, and only 19% of these facilities routinely perform testing for LB among patients at high risk for nosocomial LD. This begs the question, how does one meet the focused objective?
In the end, it can be seen that the generic system we've outlined is a superior investment at a fraction of the cost. Whatever option one chooses, it's smart to go proactive when dealing with LD.