United States Environmental Protection Agency National Risk Management Research Laboratory Research Triangle Park, NC 27711 Research and Development EPA/600/SR-95/149 April 1996 EPA Project Summary Design and Testing of Sub-Slab Depressurization for Radon Mitigation in North Florida Houses: Part I - Performance and Durability C. E. Roessler, R. Morato, R. Richards, H. Mohammed, D. E. Hintenlang, and R. A. Furman A demonstration/research project was conducted to evaluate sub-slab depressurization (SSD) techniques for radon mitigation in North-central Florida where the housing stock is primarily slab-on-grade and the sub-slab medium typically consists of native soil and sand. Objectives included developing and testing the use of a soil depressur- ization computer model as a design tool, optimization of SSD design for North Florida houses, and observation of the performance and durability of the installed systems. Between May 1989 and August 1990, SSD systems were designed and in- stalled in nine houses—seven with simple rectangular floor plans and two with more complex L-shaped designs. Installations included a single-suction- point system in one house and two- suction-point/single-fan systems in eight houses. The installation in one of the larger L-shaped houses consisted of a single-suction-point system in ad- dition to a two-suction-point/single-fan system. All systems used small diam- eter, nominal 50-mm (2-in.) piping. All houses were equipped with con- tinuous radon monitors and integrat- ing radon monitors were also deployed. All houses were visited on a regular schedule for measurements and obser- vations. The mitigation successfully reduced indoor radon concentrations originally on the order of 10 to 30 pCi/L to post- mitigation values of <4 pCi/L in all nine houses. Levels were reduced to values on the order of 2 pCi/L or less in three houses. Installation experiences demon- strated the importance of avoiding "short-circuit" air flow leakage near suction points, providing drainage for moisture that condenses in the system during cooler weather (even in Florida), and sealing around discharge ducts at roof penetrations to prevent re-entry of exhausted sub-slab gases. System manipulations indicated that a single suction point was sufficient on two houses with 160 to 170 m2 (1700 to 1800 ft2) slabs, but that passive ventila- tion is not likely to be effective for this type of sub-slab medium. During the limited time available for durability observations (3 to 18 months), the systems retained effec- tiveness in maintaining reduced indoor radon concentrations, no fans failed, and no structural effects were ob- served. This Project Summary was developed by EPA's National Risk Management Research Laboratory's Air Pollution Prevention and Control Division, Re- search Triangle Park, NC, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). Background This work was conducted in North Florida because of the presence of el- evated indoor radon1 levels and the need to investigate mitigation techniques sue- ------- cessful for the housing stock and condi- tions represented. In late 1986, results from a statewide indoor radon survey identified two focal areas of elevated indoor radon in Florida: one in the Bone Valley phosphate mining region of West Central Florida (Polk, Hillsborough, and surrounding counties); and the other in the North Florida Haw- thorn formation region—with the greatest affected populations in the Gainesville- Ocala area (Alachua and Marion coun- ties). Several other studies have confirmed the presence of indoor radon levels rang- ing from <1 to about 200 pCi/L in this area. In Florida, the housing stock is primarily of slab-on-grade construction with several variations of floor-wall joining. There is a small percentage of crawl-space and slab/ crawl-space combination houses (both open and enclosed crawl spaces), and there are very few houses with basements. The U.S. Environmental Protection Agency (EPA) has suggested that soil de- pressurization is the most successful method of limiting indoor radon; thus sub- slab depressurization (SSD) appeared to be a promising mitigation method for Florida houses. However, at the time this project was initiated, most mitigation ex- perience in the U.S. had been with base- ment houses. Furthermore, the sub-slab materials commonly used in Florida con- struction consist of native soil and sand. These would be expected to have lower air permeabilities than the coarse gravels commonly used under basement slabs in regions of the U.S. where SSD has been highly successful. This suggested that more complex and more robust systems might be required to successfully control radon in construction typical of Florida. Radon mitigation demonstration work in Florida was begun with the 1987 initiation of the EPA-sponsored Florida Radon Miti- gation Project - Phase I in Central Florida (Polk county). In late 1987, EPA also spon- sored a University of Florida (UF) project to identify elevated radon houses that might serve as candidates for a parallel North Florida mitigation project. During the 1987-88 winter, screening measurements (charcoal collector method) were made in nearly 400 Gainesville and Ocala vicinity slab-on-grade houses in neighborhoods designated on the basis of geological po- tential for elevated radon. In these screen- ing measurements on this selected group of houses, about 70% of the indoor radon concentrations exceeded 4 pCi/L and about 20% of the total exceeded 20 pCi/L. The North Florida effort continued with the August 1988 initiation of the research and demonstration project, Florida Radon Mitigation Project Phase II - North Florida. Objectives This project had a demonstration objec- tive and a series of research objectives. The demonstration objective was to dem- onstrate mitigation methods that are ef- fective for the substrate and construction type characteristic of the North Florida region. Initial emphasis was on sub-slab depressurization (SSD). The project had three research objectives: 1. Develop tools for design of SSD systems. This includes testing the use of a soil depressurization computer model2. 2. Optimize SSD design for North Florida houses. 3. Observe the short- and long-term performance and durability of the installed SSD system in this en- vironment This project involved the following work areas: 1. Select and characterize a candi- date pool of houses. 2. Mitigate a subset of these houses —select houses, design mitigation systems, and install mitigation systems. 3. Monitor: 3a. Collect baseline data prior to mitigation. 3b. Monitor initial post-mitigation performance. 3c. Conduct special studies on installed systems for the pur- pose of system optimization. 3d. Evaluate durability of in- stalled systems-continue monitoring and observations for the duration of the project. Diagnostic Methods From the data obtained in a house iden- tification study, 12 elevated radon houses were selected as potential candidates for 1The terms "radon" and "Rn" are used to designate the radon isotope, radon-222; and "radium" is used to designate radium-226. 2 Further development and testing of a computer model previously developed at UF for simulating sub-slab pressures and flows during the operation of soil de- pressurization systems was authorized. The work, which included expanding the model, developing it as an SSD design tool, and validation, is presented in Part II of this report. the mitigation demonstration. These houses were visited, and the EPA diag- nostic measurements were performed. These diagnostic observations included descriptive information, sub-slab measure- ments, radon measurements, and house dynamics observations. Sub-slab measurements included de- termination of soil gas radon by "sniff and "grab" sampling, sub-slab communication testing, calculation of effective permeabil- ity, and sampling of the sub-slab material. Indoor radon measurements included short-term measurements of concentra- tions in the living space and also mea- surements of radon in the building shell. House dynamics measurements included indoor/outdoor and indoor/sub-slab pres- sure differential measurements under vari- ous conditions and blower door pressur- ization/depressurization tests. Mitigation System Design and Installation The design procedures are described in the Part II report. Briefly, potential suction points were located on the basis of acces- sible, unobtrusive locations—usually in in- terior closets. The UF soil depressuriza- tion computer model was then used as a design tool. For an initial set of five houses selected for mitigation in 1989, the model was used to simulate pressure fields un- der proposed designs. Suction system pressures and flows were predicted by superimposing the sub-slab "system curve" on the respective fan performance curves of candidate fans. For each house, the number of suction points, their locations, and the fan size were selected from the combination giving a pressure field cover- age believed to be adequate to overcome inflow of radon-bearing soil gas. Subse- quently, the computer model was used to develop soil depressurization system guidelines for the Florida radon-resistant building code. For a second set (four houses) selected for mitigation in 1990, system designs were specified using the evolving code guidelines. To save cost, reduce space require- ments, and facilitate installation, nominal 50-mm (2-in.) polyvinyl chloride (PVC) pip- ing was specified for the major runs of the SSD systems rather than the nominal 100- mm (4-in.) piping reported in the literature for previous mitigation projects. It was an- ticipated that, because of the low flows associated with the low permeability Florida sub-slab medium, flow-related pres- sure losses in the smaller piping would not be large enough to compromise the effectiveness of the system. ------- At each suction point, sub-slab fill and/ or soil was removed to form a roughly hemispherical pit, approximately 0.5 to 0.9 m (20 to 36 in.) in diameter. Nominal 100- mm (4-in.) PVC piping with a cleanout branch to serve as an access port was installed through the slabs. The remain- der of the suction system consisted of nominal 50-mm (2-in.) PVC piping. Suc- tion piping was run vertically from the pit to the attic. Fans were located in the attic. For the systems with two suction points, lateral piping was run from the vertical risers to a tee located under the suction fan. For the single-suction-point systems, the vertical piping was run directly up to the fan. Systems were installed by the research team. Electrical hookup was provided by licensed electrical contractors. Monitoring Approach, Parameters, and Measurements At each house, monitoring was con- ducted during three time periods: 1) the baseline data collection period, 2) the sys- tem installation and tuning period, and 3) the post-installation performance/durabil- ity monitoring period. Data collection con- sisted of a combination of 1) continuous multi-parameter data acquisition (in a sub- set of four houses), 2) continuous radon monitoring, 3) integrated radon monitor- ing, and 4) point measurements and ob- servations in conjunction with site visits. The continuous recording data-acquisi- tion systems which were installed in the subset of four houses (referred to as "in- strumented houses") consisted of data log- gers with sensors for pressure differential (outdoor vs. indoor and sub-slab vs. in- door), indoor radon, temperature (indoor and outdoor), rainfall, and wind speed and direction. Data were sampled every 30 seconds and summed or averaged, and hourly sums or averages were stored in memory. Indoor radon was monitored continu- ously in all houses, either as part of the data logging system (hourly averages) in the instrumented houses or by a stand- alone continuous radon monitor (4-hour averages) in the other houses. Integrating radon monitors (electret ionization cham- bers) were also used. During site visits to the houses, pres- sures and flows were measured in the suction lines near the suction point, and "sniff and "grab" sample measurements were made of radon concentrations in the sub-slab and/or exhaust air. Qualitative observations were made of the system and house condition. Baseline Measurements Baseline data collection was targeted for at least a month-long period prior to installation of the SSD system. Measure- ments included indoor radon concentra- tion by integrating detectors, indoor radon concentration by continuous monitoring, pressure differentials (in some instru- mented houses), and weather data (in some instrumented houses). Post-Installation Performance/ Durability Monitoring Following installation and tuning of the mitigation system, continuous data acqui- sition systems (instrumented houses) or continuous radon monitors (non-instru- mented houses) were operated, integrat- ing radon monitors were deployed, and periodic house visits were performed. Post-installation monitoring was con- ducted according to the following general three-stage schedule: • Stage 1 Monitoring (in service <6 months)-Continuous and/or integrat- ing indoor radon monitoring was per- formed and houses were visited bi- weekly to observe system operation, measure pressures and flows, and service radon monitoring equipment. • Stage 2 Monitoring (in service 6 to 12 months)—Houses without data loggers were visited monthly. For houses with data loggers, data acquisition was continued, data were reviewed, and visits were performed as necessary. • Stage 3 Monitoring (in service >12 months)—As a longer-term follow-up, visits were conducted approximately every 6 months to inspect the sys- tems, measure pressures and flows, and deploy radon monitors for a week- long measurement. Performance and durability were evalu- ated in terms of: • System Performance and Interaction with the Sub-slab Medium—System pressures and flows, noise and vibra- tion, and requirements for adjustments and maintenance. • Condition of the Sub-slab Environ- ment-Effective permeability calculated from pressures and flows, and ex- haust air and/or sub-slab radon con- centrations. • Effectiveness-Indoor radon concen- trations. • Structural Effects-Observations for evidence of subsidence, heaving, cracking, separation of joints, etc. In addition, responses were made to homeowner questions or homeowner-iden- tified problems. Results and Discussion House Characterization Diagnostics were performed on 12 Gainesville and Ocala vicinity slab-on- grade houses during the last week of No- vember 1988. Installation of Demonstration SSD Mitigation Systems SSD systems were installed in nine houses: six in Gainesville and three in Ocala (Table 1). House floor plans in- clude seven rectangular and two more complex, L-shaped designs. The installa- tions include one house with a single- suction-point system, seven with two-suc- tion-point/single-fan systems, and a house with both a two-suction-point/single-fan system and a single-suction-point/single- fan system. Four houses were instru- mented for continuous data acquisition. Five of the systems were installed be- tween May and November 1989, three in Gainesville and two in Ocala. These houses were all of simple, single rectan- gular slab configuration with slab areas ranging from 158 to 195 m2 (1700 to 2100 ft2).The system at the smallest house con- sists of a single suction point and a single fan; all of the others are two-suction-point/ single-fan systems. The Gainesville houses were equipped with continuous data ac- quisition systems. During the summer of 1990, systems were installed in two additional houses with simple rectangular slabs (149 to 181 m2 or 1700 to 2100 ft2) and in two larger (195 to 203 m2 or 2100 to 2200 ft2 ) houses with L-shaped floor plans. All of these systems had two suction points con- nected to a single fan. The system in the largest house also had a third suction point with a second fan. A continuous data acquisition system was installed in one of the rectangular houses. ------- Table 1. House Summary of Mitigation Installations North Florida Project Slab, m2(ff) Operation Date Indoor Rn, pCi/L Unmitigated System on Rectangular Slabs (7 houses): Ocala-1 Ocala-2 Gainesville-1* Gainesville-2* Gainesville-3*+ Gainesville-4* Ocala-3 167 (1800) 164(1760) 164(1760) 194 (2087) 158 (1700) 181 (1950) 149 (1608) L-Shaped Slabs (2 houses): Gainesville-5# Gainesville-6 195 (2100) 203 (2188) May 1989 May 1989 Jul 1989 Nov 1989 Oct 1989 May 1990 Aug 1990 Jul 1990f Jul 1990f 16 10 11 25 9 11 30 25 26 2.5 2.0 3.5 2.5 2.0 2.6 2.0 2.5 2.5 * Continuous data acquisition systems (4 houses). f Although Gainesville-5 & -6 were turned on July 1990, they required further adjustment and became successful Oct 1990. System Types: + Gainesville-3: Single-suction-point system # Gainesville-5: Dual installation (Two-suction-point/single-fan system plus single-suction-point/single-fan system) All others: two-suction-point/single-fan systems 1 house 1 7 9 houses Mitigation Results The mitigation successfully reduced in- door radon concentrations originally on the order of 10 to 30 pCi/L to post-mitiga- tion values of < 4 pCi/L in all nine houses. Levels were reduced to values on the order of 2 pCi/L or less in three of the houses. Design and Installation Experiences Mitigation Design As indicated above, the UF soil depres- surization model was used as a design tool in placing suction points and sizing system components. The results of this work are presented in the Part II report. Moisture Condensation The early installations had some undrained low points in the horizontal pip- ing runs in the attics; with the advent of cool weather in November 1989, water condensation from the moist exhausted air essentially blocked these systems and compromised their effectiveness. This problem was overcome by installing drain lines from the moisture traps. Sub-Slab Leakage It was observed that air leakage near the suction point can compromise the sys- tem effectiveness. For example, in one case, "short-circuit" flows from a leakage crack near one suction point of a two- point, single-fan system resulted in exces- sive flows at that suction point, an imbal- ance of the system, a compromised pres- sure field, and unsatisfactory effectiveness. Caulking the crack resulted in satisfactory performance. Subsequent failure of the silicone caulking resulted in degraded per- formance; this was remediated by recaulking with urethane elastomer. Other experimental work and simulation with the computer model indicated that leakage at points more remote from the suction point has much less influence on effectiveness. Re-entrainment An adventitious experience indicated the potential for re-entrainment problems. Fol- lowing the initial installations in two houses, indoor radon levels were >:10 pCi/L when the systems were operating. Attic levels of 10's of pCi/L were found in subsequent radon monitoring. Investigation revealed that the roof penetration was not sealed around the vent pipe, apparently provid- ing the opportunity for discharged sub- slab gases to enter the attic and be drawn into the house ventilation system. Sealing the roof penetrations reduced radon con- centrations in the attics and indoors to <4 pCi/L (Table 1). Optimization Studies Pipe Sizing Nominal 50-mm (2-in.) suction piping was installed as planned. For most of the cases (61% of the suction holes), flows were sufficiently low that calculated pres- sure losses due to flow were <15 Pa/10 m, and for 90% of the holes losses were calculated to be <100 Pa/10 m. In the two cases of the highest flows where calcu- lated losses were >100 Pa/10 m, actual pressures on the order of -300 Pa (-277 to -328 Pa) were observed. The systems, involving these suction points in combina- tion with a second suction point, were effective in reducing indoor radon levels by factors of 3 to 10, resulting in indoor radon levels of 3.5 pCi/l or less for these houses. The use of the smaller piping permitted savings in cost, space, and in- stallation effort. Suction Points The effectiveness of single-suction-point operation was tested in several of houses with two-hole systems by operating these systems for a period of time with one or the other suction line valved off. These experiments indicated that: 1. Two suction points successfully maintained levels below 4 pCi/L for slab areas up to 2100 ft2. 2. A single suction point was sufficient on three houses with 1700 to 1800 ft2 slabs. Passive Ventilation The potential effectiveness of passive sub-slab ventilation was tested in several of the houses by monitoring indoor radon with the fans off and the suction lines open. These experiments indicated that passive venting (fan off, vent line open) was not effective for a packed sand/soil sub-slab medium. Durability Special questions were posed concern- ing durability for systems operating under Florida conditions. Would continued op- eration impact the sub-slab environment ------- in a manner that affects the continued effectiveness of the system? If there were effects on the sub-slab environment, would these have structural effects on the build- ing? Would continued performance of the fans be compromised by the low flow and high temperature in Florida installations? As of the end of 1990, the 1989 instal- lations had been monitored for post-miti- gation periods on the order of 13 to 19 months. Insufficient time had elapsed for significant durability monitoring on the 1990 systems which had been installed during the period May through August. During the limited observation period (3 to 18 months), the following were observed: 1. With the transient exceptions noted below, the systems exhibited rela- tively constant performance and re- tained their effectiveness in main- taining reduced indoor radon con- centrations. 2. In one case, failure of silicone caulk- ing of a leakage crack near a suc- tion point resulted in increased "short-circuit" flows. This was remediated by re-caulking with ure- thane elastomer, a more durable material. 3. With the advent of cold weather, condensation formed in horizontal attic runs that were not self-drain- ing. This resulted in an audible gur- gling noise, reduced flow, and in- creased fluctuations in indoor ra- don concentrations. This condition was remediated by installing traps and drains. 4. No fan failures were observed—any effect of low flow on fan life was not expressed during the available observation period. 5. No structural effects were observed. 6. With the exception of the "gurgling" associated with the water conden- sation before the installation of traps and drains, there were no homeowner complaints of noise or other annoyances. Conclusions and Recommendations Design Considerations 1. SSD was effective for North-central Florida slab-on-grade houses of both simple rectangle and L-shaped floor plans. 2. For the sub-slab media found in this region, low flows permitted use of smaller diameter, nominal 50 mm (2 in.) piping. 3. Two suction points were success- ful for slab areas up to 200 m2 (2100ft2). 4. A single suction point was suffi- cient on three houses with single- level, rectangular slabs with areas on the order of 160 to 170 m2 (1700 to 1800ft2). 5. Experiments with installed active systems (fan off, vent line open) indicated that passive ventilation is not likely to be effective for this type of sub-slab medium. Installation Considerations 1. Air leakage near the suction point can compromise system effective- ness; leakage at points farther from the suction point has much less influence on effectiveness. 2. Even in Florida, moisture can con- dense in the system during cooler weather; it is important to avoid low points in horizontal attic runs and to install traps and drains if water trapping points cannot be avoided. 3. It is important to seal around the discharge duct at the roof penetra- tion to prevent re-entry of the ex- hausted sub-slab gases. Examina- tion for other sources of re-entrain- ment is also warranted. Performance and Durability The following conclusions are limited by being based on short observation times- 3 to 18 months: 1. Pressure and flow values in SSD systems may exhibit some tempo- ral variability; documentation of per- formance from point measurements should be based on averages from a series of measurements taken on different days. 2. On a near-term basis, SSD sys- tems as installed in this project re- tain effectiveness in maintaining re- duced indoor radon concentrations. 3. Continued integrity of sealing of po- tential short-circuit air flow sources near suction points is essential to continued effectiveness. System maintenance should include inspec- tion of such sealing. 4. During cooler weather, unintended trapping of moisture condensation in horizontal attic runs can compro- mise system performance. Mainte- nance should include inspection for such inadvertent effects. 5. Fan failures have not been identi- fied as a problem in the short term (based on observing a small num- ber of systems). 6. Structural effects have not been identified in the short term. 7. Other than for the noises associ- ated with the water condensation before correction, these systems have not generated homeowner complaints. Short-term durability information would be enhanced by following all houses for at least a year, and long-term durability in- formation would be gained by following all the houses even longer. ------- C. Roessler, R. Morato, R. Richards, H. Mohammed, D. Hintenlang, and R. Furman are with the University of Florida, Gainesville, FL 32611. David C. Sanchez is the EPA Project Officer (see below). The complete report consists of two volumes entitled "Design and Testing of Sub- Slab Depressurization for Radon Mitigation in North Florida Houses: Part I. Performance and Durability." "Volume I. Technical Report" (Order No. PB96 -103 585; Cost: $21.50, subject to change) "Volume II. Data Appendices," (Order No. PB96-103 593; Cost: $35.00, subject to change) The above reports will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air Pollution Prevention and Control Division National Risk Management Research Laboratory U. S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency National Risk Management Research Laboratory (G-72) Cincinnati, OH 45268 Official Business Penalty for Private Use $300 BULK RATE POSTAGE & FEES PAID EPA PERMIT NO. G-35 EPA/600/SR-95/149 ------- |