5.1 Granular Media Filtration Theory
Key Takeaways
- Rapid sand, dual/multimedia, and slow sand filters differ in media arrangement, rate, and cleaning; dual/multimedia enable coarse-to-fine depth filtration.
- Effective size (d₁₀) and uniformity coefficient (d₆₀/d₁₀) describe media grading and strongly influence headloss and filtration behavior.
- Particle removal uses straining, sedimentation, interception, and adsorption—proper coagulation is required for attachment, not sieving alone.
- Operators track headloss versus turbidity breakthrough and use filter-to-waste after backwash during ripening before returning a filter to service.
- LT1/LT2 and related surface-water rules frame filtration turbidity performance as a treatment technique that supports pathogen log-removal credit.
5.1 Granular Media Filtration Theory
Quick Answer: Granular media filtration removes suspended particles after coagulation/sedimentation (or as direct filtration) using beds of sand, anthracite, garnet, or multimedia layers. Removal mechanisms are straining, sedimentation, interception, and adsorption—not a simple sieve. Operators manage media size and uniformity coefficient, track headloss and turbidity breakthrough, use filter-to-waste after backwash, and understand that Surface Water Treatment Rule / LT1 / LT2 frameworks award pathogen log-removal credit when filter performance meets turbidity treatment-technique criteria.
Filtration is a core barrier on conventional and direct-filtration plants and on many groundwater systems that oxidize and filter iron and manganese. FDEP Class C and Class B drinking-water outlines expect you to know filter types, media characteristics, particle-removal mechanisms, and how turbidity and headloss signal a successful (or failing) filter run. Exam items often ask what is happening inside the bed when effluent turbidity rises, or which media change improves run length without sacrificing quality.
Why Filtration Matters in the Multi-Barrier Chain
Particle removal protects public health because:
- Giardia cysts and Cryptosporidium oocysts are relatively resistant to free chlorine; physical removal is essential.
- Turbidity shields pathogens from disinfectant contact and raises chemical demand.
- Well-operated filters produce low effluent turbidity that supports disinfection CT and distribution water quality.
Filtration is usually the last major solids-removal step before disinfection. Coagulation and flocculation prepare particles; sedimentation (on conventional plants) reduces load; the filter polishes residual floc and fine colloids.
Rapid Sand vs Dual/Multimedia vs Slow Sand
| Filter type | Typical media | Filtration rate (order of magnitude) | Pretreatment | Exam takeaway |
|---|---|---|---|---|
| Rapid sand | Graded silica sand (single media) | Roughly 2–5 gpm/ft² (plant-specific) | Coagulation; often sedimentation | Deep-bed filtration with frequent backwash; not a surface strainer only |
| Dual media | Anthracite over sand | Similar range; often higher than mono-sand for same quality | Coag/floc ± settling | Coarse-to-fine depth filtration; longer runs, better use of bed |
| Multimedia (mixed media) | Anthracite / sand / garnet (top to bottom by density) | Comparable to dual; design-specific | Coag/floc ± settling | Density-graded layers stay stratified after backwash; maximum bed utilization |
| Slow sand | Fine sand, biological schmutzdecke | Roughly 0.05–0.15 gpm/ft² | Limited; often no coagulant | Biological + physical removal; scrapes/replaces surface, not high-rate backwash like rapid filters |
| Pressure filters | Sand or dual media in closed vessels | Similar to rapid rates | Common for Fe/Mn, smaller plants | Same theory; monitor differential pressure as headloss proxy |
Rapid Sand Filtration
Rapid sand filters use relatively coarse sand compared with slow sand and are backwashed with reverse flow (often with air scour). After backwash, media can stratify by size with finer grains on top—that is the opposite of ideal depth filtration and is one reason dual/multimedia designs were developed. Rapid filters depend on well-conditioned floc; they are not designed to remove large raw-water solids without coagulation.
Dual and Multimedia Filtration
Dual-media beds place anthracite (larger, lower density) over sand. Coarse upper media traps large floc first; finer sand below polishes. That coarse-to-fine path uses more of the bed depth, delays terminal headloss, and often improves effluent quality.
Multimedia adds a dense fine layer (often garnet) at the bottom. After backwash, layers re-stratify by density: anthracite on top, sand middle, garnet bottom—preserving coarse-to-fine filtration. Exam language: depth filtration versus surface straining.
Slow Sand Filtration
Slow sand develops a biological schmutzdecke on the surface that contributes to pathogen and organic removal. Rates are much lower; cleaning is typically scraping the surface layer rather than hydraulic backwash cycles used on rapid filters. Slow sand is less common at large modern Florida surface plants than rapid dual-media designs, but the contrast appears on exams: biological surface mat vs engineered deep-bed rapid filters.
Media Sizes and Uniformity Coefficient
Operators and designers specify media with two classic metrics:
Effective Size (ES or d₁₀)
Effective size is the sieve opening that passes 10% by weight of the media sample. Smaller effective size generally means finer media, tighter pores, better particle capture at the cost of higher headloss and sometimes shorter runs if load is high.
Typical order-of-magnitude values operators memorize for exams (plant specs govern real media):
| Media | Typical effective size range (approx.) | Role |
|---|---|---|
| Anthracite | ~0.8–1.2 mm (varies) | Upper coarse layer |
| Filter sand | ~0.45–0.55 mm common design band | Main filtration media |
| Garnet | ~0.2–0.4 mm (finer) | Bottom polishing layer |
Uniformity Coefficient (UC)
Uniformity coefficient = (d_{60} / d_{10}) (size that passes 60% divided by effective size).
- Lower UC (closer to 1) means more uniform grain size.
- Very uniform media can give more consistent porosity; excessively wide size distribution can pack tightly, raise headloss, and behave poorly in backwash.
- Design specs often target UC values roughly in the 1.3–1.7 band for many filter sands (know the definition and the qualitative effect, not a single universal number).
Why exams care: if media is worn, broken, or contaminated with mudballs, effective size and UC in practice change—hydraulic behavior and water quality suffer even if the original design was correct.
Particle Removal Mechanisms (Not Just a Sieve)
Particles smaller than pore openings are still removed. Four mechanisms dominate teaching materials:
- Straining — particles larger than pore openings are trapped at grain contacts or the bed surface. Important for large floc, but not the only mechanism for fine colloids.
- Sedimentation — within pore spaces, particles settle onto media grains under gravity (pore “mini-clarifiers”).
- Interception — particles following streamlines contact a grain and stick when the streamline passes within about one particle radius of the surface.
- Adsorption / attachment — van der Waals and chemical forces hold particles to grain surfaces after transport; proper coagulation is critical because poorly neutralized particles may collide but not attach (or detach later).
Additional exam-relevant ideas:
- Diffusion (Brownian motion) matters more for very small colloids and some microbes.
- Biological activity in slow sand (and some biofilters) adds biodegradation beyond pure physical capture.
- Ripening after backwash: clean media needs a short period (and often filter-to-waste) before attachment sites and residual floc improve removal efficiency.
If coagulation fails, filters can show early turbidity breakthrough even when headloss is still low—particles are not sticking.
Headloss and Turbidity Breakthrough
A filter run is a balance between water quality and hydraulics.
Headloss
Headloss is the pressure (or water-level) drop across the media caused by friction and clogging. As solids accumulate:
- Clean-bed headloss starts at a baseline set by rate, media, and temperature.
- Headloss rises through the run as pores fill.
- Terminal headloss is the design or operational limit that triggers backwash (along with time or turbidity rules).
Operators watch rate-of-headloss-rise. A sudden jump can mean surface blinding (poor pretreatment, algae, polymer overdose) or hydraulic issues. A very slow rise with rising turbidity can mean media problems or inadequate floc strength (particles pass through without building headloss).
Turbidity Breakthrough
Breakthrough is rising filter effluent turbidity while the bed still has hydraulic capacity—or at end of run when storage is exhausted. Causes include:
- Incomplete coagulation / weak floc
- Hydraulic surge or abrupt rate increase
- Mudballs, media cracks, or short-circuiting
- Excessive run length
- Air binding or media loss exposing support gravel
Rule of thumb for operations: backwash on the first limit reached—turbidity, headloss, or maximum run time—whichever plant SOP and regulatory strategy require.
| Signal | Likely interpretation | Operator direction |
|---|---|---|
| High headloss, low turbidity | Solids captured well; bed loaded | Plan backwash; pretreatment OK |
| Low headloss, rising turbidity | Poor capture / short-circuiting / weak floc | Fix chemistry; inspect media; do not “run longer” |
| Both rising together near end of run | Normal mature run | Backwash; review if run is too short |
| Instant turbidity spike after rate change | Hydraulic shear / detachment | Avoid abrupt rate changes; ramp flow |
Filter-to-Waste
After backwash, effluent quality is often worse for minutes because:
- Residual backwash water is turbid
- Media is expanded and not fully settled
- Attachment is inefficient until ripening occurs
Filter-to-waste (FTW) sends initial filtered water to waste (or reclaim) until online turbidity meets return-to-service criteria. Many plants use continuous effluent turbidimeters and automatic FTW valves. Exam point: returning a filter to service immediately after backwash without ripening/FTW risks a turbidity spike to the clearwell and can compromise treatment-technique performance.
LT1 / LT2 Pathogen Log-Removal Credit (High Level)
Federal surface-water rules (implemented through state primacy programs such as Florida’s FAC drinking-water framework) treat filtration performance as a treatment technique for pathogens, not merely an aesthetic goal.
At the conceptual level operators need for exams:
- Surface Water Treatment Rule (SWTR) family requirements apply to surface water and GWUDI sources.
- Well-operated conventional or direct filtration trains are awarded log-removal credits for Giardia (and virus credit is largely from disinfection CT; Cryptosporidium credit depends on rule tier and bin).
- LT1ESWTR extended strengthened requirements to smaller systems; LT2ESWTR requires Cryptosporidium source monitoring for many surface systems and additional treatment based on bin classification.
- Combined filter effluent (CFE) turbidity criteria historically emphasized by operators—commonly taught as ≤ 0.3 NTU in at least 95% of monthly samples and never exceeding 1.0 NTU for conventional/direct filtration under the strengthened rules—are the performance surrogate that supports those credits.
- Individual filter effluent (IFE) monitoring catches a single bad filter that CFE averaging might hide.
You are not expected to recite every EPA bin table from memory on every Class C item, but you must connect: good granular filtration turbidity performance = pathogen removal credit; failed turbidity performance = public-health and compliance risk, not just a “dirty water” complaint.
Florida Operating Context
Many large Florida surface and GWUDI plants use dual-media rapid filters after conventional pretreatment. Groundwater plants may use pressure filters after aeration/oxidation for iron and manganese—same attachment/straining theory, different particle chemistry. Coastal and inland systems that later add membranes still rely on granular pretreatment in hybrid trains. Whatever the configuration, the exam tests whether you understand depth filtration, media specs, headloss vs breakthrough, filter-to-waste, and turbidity-based pathogen credit.
Operator Memory Hooks
- Dual/multimedia = coarse-to-fine depth filtration.
- UC = (d_{60}/d_{10}); ES = (d_{10}).
- Mechanisms: strain, settle, intercept, adsorb—coagulation makes adsorption work.
- Backwash on turbidity or headloss or time—do not ignore early breakthrough.
- FTW after backwash until ripening meets turbidity return criteria.
- LT1/LT2: filtration log credit depends on meeting turbidity treatment technique, not on “looking clear.”
A dual-media filter places anthracite over sand. What is the primary operational advantage of this arrangement compared with a single sand bed of uniform fine media?
Uniformity coefficient (UC) of filter media is defined as which ratio?
Filter effluent turbidity begins rising while headloss is still low midway through a run. What does this pattern most strongly suggest?
Why do plants use filter-to-waste immediately after backwash before returning a filter to the clearwell?