Attendees: James Jones (DC Appleseed), Paul Schwartz (Clean Water Action), Wayne Turner (DC Appleseed), Lloyd Stowe (Washington Aqueduct), Ted Tulle (District Department of the Environment), Ralph Scott (Alliance for Healthy Housing), George Hawkins (District Department of the Environment), Maureen McGowan (Office of the City Administrator/DDOE), Yanna Lambrinidou (Parents for Non-Toxic Alternatives), Dina Passman (Department of Health), Chevelle Glymph (Department of Health), Maureen Schmelling (DC WASA), Rich Giani (DC WASA), Joe Cotruvo (DC WASA Board), Charlie Kiely (DC WASA), Tom Jacobus (Washington Aqueduct), Jeff Kempic (US EPA). Lisa Donahue (US EPA) and Dr. Marc Edwards (Virginia Tech) participated via phone. 

 

The meeting opened with a spreadsheet presentation summarizing Task Force (TF) members’ opinions on how to structure an eventual study.  (Note: Those issues discussed from the spreadsheet appear as headings below).  Mr. Scott requested clarification of the suggestion that lead and iron both be included.  Mr. Kiely explained that DC WASA has been investigating whether a potential correlation between the two exists.  Since copper levels have always registered as low, it may be worthwhile to study iron instead. 

 

Scope of the initial RFP
Mr. Cotruvo asked whether the initial RFP will cover the entire scope of contaminants or just lead.  Mr. Hawkins stated that his sense is that the study will cover just lead initially, so that the TF doesn’t lose too much time before the summer season.  The eventual straw proposal will allow the TF jurisdiction over a broader range of issues that may be included in a later round of testing.  It is a reasonable suggestion to cover iron.  Ms. McGowan suggested writing the RFP with a supplemental option, such that the main scope is set, but if bidders can perform additional services, they may propose to do so in their submission. 

 

Mr. Cotruvo suggested that it might be doable to lump all of the contaminant testing together.  Organic chemical side of things is much easier to carry out—the labs would perform a cross-sectional assessment of what is in the water. Ms. Donahue advised the TF to think about what other parameters need to be collected at the same time—temperature, disinfectant residual, etc.  Some of those elements might be helpful in answering both lead questions and informing future pieces of the study.  Mr. Scott asked whether some of that information might be available from the Aqueduct, given that there might be some difficulty in collecting it reliably via individual samplers.  Ms. Donahue responded that while some of the information can be found system-wide, EPA has found that “if we only had information about associated parameters, we might have had a clearer picture.”  Since testing never can capture exact conditions again, it’s best to get the most exact picture of the water on that day.  If homeowners are taking the samples, some of the parameters could potentially be determined once they reach the lab, so that testing is more practical. 

Mr. Kiely stated that 87 out of 199 samples were returned during recent LCR testing. Therefore, the TF must consider the likelihood if that the homeowner would be available if the same pool of homes was re-sampled.
Mr. Hawkins explained that the instinct is to do both so that we may compare the difference between citizen-collected samples and highly controlled, technician-collected samples.  He also posited that on questions of statistical validity that might be posed by a smaller sample set, an early smaller phase might just be used to show outcomes that could be investigated further in a later phase of the study.  Ms. Passman suggested that a literature review to support some of these decisions would be helpful.  Mr. Schwartz posed the question of what the TF knows about different sampling protocols and methodologies.

 

Mr. Scott suggested that if technicians come early, they should also collect samples at a later hour to reflect the times when people are more likely using the water.  Dr. Edwards pointed out that sampling with technicians raises costs significantly—there’s a reason that utilities rely on customers.   He disagreed with the general idea that there’s a certain professionalism involved in sample collection.  The LCR currently allows homeowners to collect samples and they generally do a good job of collecting. 

 

Where should sampling be done?
Mr. Kiely started the discussion by proposing that it is a good idea to test schools, but would probably require a different protocol, since residential service testing is different from that of testing a large building. Dr. Edwards suggested using the EPA lead in school protocol. This options uses smaller sample volumes and the second draw samples tells you less, but once can still learn a lot from samples at individual taps. 

Dr. Edwards posed a second option—that many TF participants want to see LCR compliance sampling done using all of the criteria for site selection, focusing on at-risk homes, and following the intent of the LCR protocol.  All other studies sound fascinating, but the main focus should be LCR testing focused on the highest risk homes in the EPA pool.  Ms. Lambrinidou added that the eventual Request for Proposals should put duplication of LCR as the first phase, and then ask for additional bids/information on additional sampling pieces/questions.  Mr. Cotruvo agreed that following the LCR is the lowest cost option.  He laid out two options: to provide quality control oversight on what WASA usually does or to do a parallel study.  Key questions include: Are the sampling locations appropriate?  Is the data valid?  According to Mr. Cotruvo, the simplest option would be a very careful quality control on WASA’s six-month set of samples. 

 

Ms. Glymph noted that when sampling sub-populations, there’s not a lot of bang for your buck if you do all of the ones proposed.  What group is the TF really interested in?  There’s always the option to say what the great next steps would be once we have the initial data. 

 

Mr. Scott believes that limiting the TF to the pool that WASA is already testing is a foolish idea.  His biggest suspicion is that WASA is picking a pool where they know they’ll get compliance.  According to Mr. Scott, in past years, turnover rates within the sampling pool have been 86%, 75%, and 63-67%.  The last thing he would agree to is to test only the homes that WASA has already been testing. 

 

Mr. Jones cautioned that the TF can get too hung up on the LCR—what he really wants is to look at the highest risk homes. This process isn’t about compliance at all; it’s about the health risks to people from lead.  Mr. Stowe pointed out that the LCR wasn’t designed to get the worst case lead levels.  Mr. Fellows countered that the intent of the LCR is to get at the worst case scenarios.  He strongly questions whether WASA did capture the worst-case.  Just duplicating WASA’s sampling wouldn’t make sense. 

 

Mr. Kiely stated that the main question is whether the water is safe to drink—not for certain populations but for everyone.  The District Inspector General issued a report that mirrored WASA’s report.  Mr. Kiely felt the lead service line inventory should be the sample population, with the TF informing how the worst-cases are selected. 

 

Mr. Kempic stated that according to the LCR, the sites most at-risk are going to be lead service lines.  Other sites were those built after lead service lines were used.  In the District, we know which sites are going to present the highest risk.  Dr. Edwards added that WASA has information beyond lead service line replacement—WASA has test results for a set of homes.  They also have information on water use rates per home, results from 2004. Dr. Edwards suggested putting this information into a database and then trying to select the worst case scenarios.  Ms. Donahue cautioned against using data that pre-dates the change in disinfection that the Aqueduct instituted after 2004 because that resulted in a significant shift.  Those homes are now under different water quality conditions. 

 

Mr. Hawkins clarified that to answer the question is whether the water is safe, the TF is not limited to sites that WASA has done-there can be an independent study with our own selection of sites. 

 

Mr. Schwartz added that another piece on the table is to instill confidence. History does matter to inform how we move forward.  Mr. Scott stated that WASA has data on homes on lead service lines.  They know which ones are testing low.  Mr. Scott would like to see information on past lead testing to see which had high results and thus are high risk.  Dr. Edwards stated that the 2004 data would give a picture of which might have problems today.  Those that tested low in 2004 without the water inhibitor are unlikely to have problems. 

 

Mr. Cotruvo suggested that if purpose is to identify worst-case homes, the TF could look at the last 2 years of data.  If the question is to do an inventory to find homes with high values, that requires a different approach.  If the question is to find homes with a health risk, the TF can’t do that at all with any of the measures that we’re talking about.  Instead, the TF would have to focus on exposure, blood lead, is it related to the water, what level of harm is produced.

 

Ms. Lambrinidou stated that the goal is simple-to duplicate LCR, including its limitations, but this is what every water utility does to measure water quality.  She believes that how the LCR has been applied has been unsatisfactory. Mr. Scott added that since LCR is designed to get at worst-case scenario, he believes it will enable us to get at health risks.  Dr. Edwards asked whether WASA’s compliance depend on what homes are sampled within the pool? He maintained that WASA selects certain homes.

 

Mr. Hawkins summarized the discussion as two main sampling options emerging: random selection from within the lead inventory and finding worst-case scenario homes using specified selection criteria.

Mr. Kempic asked how you select worst sites within the lead service line inventory. There is a debate over how to define highest risk within that predetermined category.  Ms. McGowan explained that they are not necessarily mutually exclusive- the TF could select homes from within a pool of homes with high test results and also select a random group of homes. 

 

Dr. Edwards stated that there isn’t a mechanism that site selection is random.  Mr. Giani explained that WASA took a computer to select 500 random sites out of 15,000 homes in lead service line.  This was done in 2004-5.  In 2003, the list was very small and designed for reduced monitoring (which WASA qualified for at the time).  A lot of initial homes on that list didn’t meet EPA requirements. 

 

Dr. Edwards asked whether within the pool, the EPA or LCR specifies that sites are random.  Mr. Kempic answered that the EPA only determines the criteria that homes must meet.  Mr. Cotruvo suggested that the TF should duplicate the LCR, but other studies could be done in parallel.  The only conclusions would be on specific homes. 

 

Dr. Edwards reminded the TF that non-compliance with the LCR would trigger public education.  Mr. Schwarz added that despite the LCR being about corrosion control, it is widely viewed as a standard that people can look to in order to learn about the state of the water.  If other aspects can be added on, that would be great, but the LCR should be the de minimis. 

 

Mr. Cotruvo posed the following questions -  Should the TF look at the quality control over existing WASA process? Are there homes with higher lead? What factors in the homes/water are contributing?  Ms. Glymph stated that whether there are three pools or two pools, it makes sense to pick the same criteria across all three pools.  This helps assist with feeling like it’s not an exact duplication, addresses concerns about statistical significance, increases validity. Pools themselves should be held to the same criteria. 

Mr. Hawkins summarized thusly: That the direction of the TF would be towards the selection of different pools (random, WASA existing, worst case according to specified criteria), with testing of all three.