Sand bioreactor design is very simple. Three major factors must be considered in the design: media depth, media characteristics, and area loading rate. Sand depth is the first design consideration. In general, the deeper the bioreactor, the greater the level of treatment. However, most of the treatment occurs in the top 9 to 12 inches. Deeper filters produce a more consistent quality effluent, but after 24 inches no significant treatment improvement is achieved with added depth. Additionally, air penetration into deep sand beds is more difficult, so deeper bioreactors are more likely to clog. Therefore, a sand depth of 24 inches is appropriate for most domestic wastewaters.
Of all of the design criteria, media characteristics are the most important. Sand bioreactor clogging is usually the result of using sand that is too fine, has too many fines, or has a weak or platey structure. The most important feature of the sand is not the grains, but rather the pores the sand creates. The treatment of wastewater occurs in the pores, where suspended solids are trapped, microorganisms grow, and air and water flow.
The ideal media is hard and nearly spherical in shape. Quartz sand is often used because it is inexpensive and readily available. Garnet sand, mineral tailings, expanded clays, and other materials have all been successfully used. Some plastic media may also be appropriate.
The size distribution of the sand is measured as the effective size and the uniformity coefficient. Ideal sands for intermittent bioreactors are a medium to coarse sand with an effective size between 0.3 mm and 1.5 mm. The uniformity coefficient should be less than 4.0. A summary of sand effective size and uniformity coefficients by bioreactor type is listed in Table 1.
| Table 1. Basic Sand Bioreactor Design Criteria. | |||
|---|---|---|---|
| Bioreactor Type and Performance | Effective Size | Uniformity Coefficient | Daily Area Loading Rate of Primary Treated Wastewater |
| Intermittent (Single pass) | |||
| Very low effluent CBOD5 and ammonia (for stream discharge) | 0.3-0.5 mm | less than 4 | up to 1 gal/ft2/day |
| Low effluent CBOD5 (for irrigation on public access sites) | 0.5-1.0 mm | less than 4 | up to 1 gal/ft2/day |
Long operation without resting (for buried bioreactors) | 0.5-1.5 mm | less than 3 | up to 0.5 gal/ft2/day |
|
Recirculating | |||
| Very low effluent CBOD5 and ammonia (for stream discharge) | 0.3-0.5 mm | less than 4 | up to 5 gal/ft2/day |
Low effluent CBOD5 (for irrigation on public access sites) | 0.5-1.5 mm | less than 4 | up to 5 gal/ft2/day |
Long operation without resting | 1.0-1.5 mm | less than 3 | up to 3 gal/ft2/day |
Sand effective size and uniformity coefficient affect filter performance. BOD5 and ammonia removal are a function of effective size. If the bioreactor effluent will be discharged to a stream, very low CBOD5 (10 mg/l) and ammonia (1 mg/l in summer and 3 mg/l in winter) effluent concentrations are typically required. Bioreactors for this purpose should be constructed of sand with effective size between 0.3 and 0.5 mm. Clogging becomes a major concern when using sand with an effective size between 0.3 mm and 0.5 mm, therefore filters using this size sand must be lightly loaded or rested periodically.
If the effluent will be reused through irrigation, required CBOD5 levels are less stringent (25 mg/l) and ammonia removal is not necessary. Sand effective size of 0.5 mm to 1.0 mm for single pass bioreactors, and 0.5 mm to 1.5 mm for recirculating bioreactors are appropriate. Clogging is less of a concern when using coarser sand.
The sand uniformity coefficient will have an effect on the time of clogging or longevity of the bioreactor. Bioreactors constructed of sand with high uniformity coefficients will begin ponding more quickly and require more frequent resting. If a bioreactor must operate for many years without resting, use sands with a uniformity coefficient less than 3. If periodic resting is planned, sand with a uniformity coefficient of up to 4 can be used. The availability and cost of the sand will greatly impact the decision. Sands with uniformity coefficients between 3 and 4 are more readily available and are lower in cost than sands with a uniformity coefficient less than 3. If sand with a uniformity coefficient between 3 and 4 is used, simply make provisions for multiple beds with periodic resting capability.
The effective size is defined as the sand size, when no more than 10% by weight is smaller, referred to as D10. The uniformity coefficient is the ratio of the sand size when 60% by weight is smaller, or the D60, over the size when no more that 10% by weight is smaller, the D10. Determining the effective size and uniformity coefficient is accomplished by screening the sand through a series of sieves, as described in Box 1.
BOX 1Determination of Sand Effective Size and Uniformity CoefficientApparatus
Begin with about a 100-gram sample of sand. Dry in 105-110oC oven for two hours. Weigh dry sand sample (WD). Label and weigh metal sample pans, and set aside. Fill sand sample container with tap water, shake and decant wash water through No. 200 sieve. Wash material retained on sieve back into sample container. Repeat several times until wash water is clear. Dry sand again in 105-110oC oven for two hours. Weigh dry washed sand (WDS) and subtract from dry weight to determine weight of fines. Wt. of fines = WD - WDS Arrange a set of sieves from largest opening to smallest as shown in Figure 7. Shake stacked sieves, vibrating, jogging, and jolting them to keep the sand in continuous motion for two minutes. Shake each sieve individually over a clean tray to make sure all the sand has passed through and is distributed by size. Pour the sand off each sieve into labeled, weighed pans. Weigh and determine the sample weight (WS) by subtracting the weight of the pan. Determine the percent passing for each sieve by: Percent of material retained on the sieve = WS x 100% / WDS Percent passing = percent passing the next largest sieve - percent retained on sieve An example calculation of percent passing each sieve for a 120 gram sample is summarized in Table 2. Graph the percent passing results on semilog paper as shown in Figure 14. From the graph, find the Effective Size as D10, where only 10% of the sample is a smaller size. Also from the graph, find D60, where 60% of the sample is a smaller size. The Uniformity Coefficient is D60/D10. |
Figure 7. Sieves
with various sized openings are used for sand analysis. The sieves are
arranged largest to smallest from top to bottom, as shown.
| Table 2. Sand Particle Size Analysis-Calculating Percent Passing Selected Sieves. | |||||
|---|---|---|---|---|---|
|
Sieve number | Sieve size | Sample weight | % passing next larger sieve | % retained | % passing |
| 3.5 | 5.60 | 6.00 | 100 | 5 | 95 |
| 10 | 2.00 | 8.40 | 95 | 7 | 88 |
| 20 | 0.85 | 57.60 | 88 | 48 | 40 |
| 30 | 0.425 | 14.40 | 40 | 12 | 28 |
| 40 | 0.425 | 12.00 | 28 | 10 | 18 |
| 60 | 0.25 | 15.60 | 18 | 13 | 5 |
| pan | -- | 6.00 | 5 | 5 | -- |
| Total Sample Weight | -- | 120.00 | |||
The area loading rate of the bioreactor is the third design criteria. In general, the higher the area loading rate, the more likely the bioreactor is to clog and back up. Once clogged, the bioreactor must be rested for four to six months.
Design area loading rates differ with filter type. A typical design area loading rate for an intermittent sand bioreactor of 1 gal/ft2/day balances the organic loading from domestic sewage to the degradation activities of the microorganisms. Bioreactors loaded at this rate can operate for many years with no ponding. Because access and air penetration is restricted in buried sand bioreactors, a very low area loading rate of 0.5 gal/ft2/day is recommended.
To save space, some intermittent sand bioreactors are loaded up to 5 gal/ft2/day. When loading bioreactors at this high rate, clogging can occur in as quickly as one to two years. Therefore, it is necessary to construct two or more bioreactor cells and carefully manage the whole system through alternating loading onto cells throughout the year. One management strategy is to divide the bioreactor into four cells as shown in Figure 6. During the warmest months of the year, rest one cell beginning in the spring (March through June) and one in the fall (July through October), while temporarily increasing the area loading rate to the other three. Continue to rest each of the other cells in the next year to complete a two-year management schedule. In this way, the per cell area loading rate is lowest in the coldest time of the year when the biological activity is also at its lowest.
Another way to reduce the area of the bioreactor is through recirculation. From 3 to 5 gallons/ft2/day of primary treated wastewater is mixed with 9 to 25 gallons/ft2/day of sand bioreactor effluent. The resulting mixture is applied throughout the day in small doses totaling 12 to 30 gallons/ft2/day.