Denitrification Biofilters: How DNBF Media and Loading Decide Whether Class 1A Nitrogen Is Reachable

Ask a plant operator what limits total nitrogen on a municipal upgrade and the answer usually lands on biology, not filtration. Yet the unit that actually decides whether Class 1A or quasi-Class IV is met on any given day is often the denitrification filter sitting at the very end of the train. It does two jobs at once: it feeds carbon to denitrifying bacteria under anoxic conditions so nitrate nitrogen (NO3-N) and nitrite nitrogen are reduced to nitrogen gas, and it physically strains out suspended solids on the way through. The first unit of this kind was commissioned in 1969, and several hundred have run worldwide since.

Where the filter sits, and why that choice matters

Placement is a design decision, not a formality. Set after secondary biological treatment, the filter works as a post-denitrification step for plants where nitrification is already complete. Coupled in series ahead of a nitrification filter, it becomes a pre-denitrification stage, a layout that suits smaller facilities where every metre of tankage is contested. Geology, raw water quality and how much attention the operations team can realistically give the unit all feed into the choice. Get the position wrong and no amount of downstream tuning recovers the lost nitrogen removal.

Media: the layer that sets the performance ceiling

Most denitrifying deep-bed filters run on 2-3 mm quartz sand, with a bed depth around 1.83 m. That depth is not arbitrary. It is what lets effluent SS stay below 5 mg/L without chemical polishing, and it is what holds TP under 0.3 mg/L once a phosphorus precipitant is dosed. Ceramsite is the common alternative. Work comparing quartz sand against biological ceramsite with methanol and sodium acetate as carbon sources found ceramsite carrying more biomass and delivering better pollutant removal, though methanol gave the more stable effluent with almost no nitrite accumulation.

Designers typically size to an empirical hydraulic load of 0.5-3 m3.m-2.h-1. Push past that and the filter stops behaving like a filter: contact time collapses, biomass is scoured faster than it regrows, and nitrate slips through with the solids.

Keeping the biology alive

Denitrifying bacteria want a narrow window: pH 6-8, water temperature 20-35 degrees C, and dissolved oxygen held at or below 0.5 mg/L. Oxygen is the usual culprit when a filter underperforms, because it suppresses the reductase enzymes the whole process depends on. In practice the carbon-nitrogen ratio is held at or above 5:1, with carbon source type and dose, hydraulic retention time (HRT), DO, pH and temperature together determining whether the biology holds. Start-up differs by carbon source too: sodium acetate brought one filter to stable operation in 7 days, while methanol needed 9.

Carbon selection is rarely purely technical. Methanol tends to give steadier effluent and less nitrite accumulation and is often cheaper per unit of nitrogen removed, but it carries handling and safety obligations. Sodium acetate responds faster and is simpler to store, at a higher unit cost. Most plants end up deciding on water quality, safety rules and delivered price together rather than on denitrification kinetics alone.

Up-flow or down-flow

Up-flow units pass wastewater from the bottom upward through the distribution layer, support layer, filter layer and clear water layer. Down-flow units feed from a distribution tank at the top and stack the distribution zone, filter zone, support layer and effluent collection zone vertically. Up-flow has the edge on denitrification efficiency and phosphorus removal, consumes less carbon, and its air-and-water distribution design relieves media clogging. Down-flow bears a structural resemblance to the V-shaped filter many teams already know how to run, which counts for something when hiring and training are tight.

Backwash: the 4% budget

No filter runs indefinitely without cleaning. Combined air and water backwash restores head loss and strips aged biofilm, and the spent washwater normally returns to the head of the plant. That recycle usually stays under 4% of total plant throughput — small enough to ignore in the water balance on a normal day, large enough to matter during peak loading or when several filters backwash in sequence. Newer installations use optimised filter bricks for air and water distribution, which evens out the backwash and shortens it, and intelligent backwash triggering on head loss rather than a fixed clock is now standard on well-instrumented plants.

Autotrophic routes and what comes next

The dependence on bought-in carbon is the weak point of heterotrophic denitrification: dose too much and effluent COD and BOD climb, which is a compliance problem in its own right. Autotrophic denitrification filters sidestep this by using inorganic electron donors such as sulfur and iron, adding no organic carbon and generating far less sludge. Sulfur-based composite carriers have reached total nitrogen removal of 90% or better in full-scale work, and the route suits low carbon-nitrogen ratio wastewater particularly well.

Three families are in common use: the denitrifying deep-bed filter developed from the traditional V-shaped filter, the activated sand filter that folds coagulation, clarification and filtration into one shell, and the denitrification biofilter (DNBF), an evolution of the biological aerated filter (BAF). Active research is now shifting toward hybrid media — biomass-iron-based beds, for instance — aimed at removing nitrogen and trace antibiotics in the same pass. In China, installed capacity using denitrification filters already exceeds one million tonnes per day, so incremental gains in carbon efficiency translate into very real operating money.