Combining RDF, Moving Bed and Bakki Shower Filtration

The Short Answer

RDF = remove solid waste

Moving bed = convert dissolved nitrogen waste

Bakki shower = additional biological treatment + intensive air–water contact

The RDF should normally come before the biological filtration because rapid solids removal reduces the organic burden placed on the biofilters.

Research in recirculating aquaculture systems consistently identifies rapid particulate-waste removal as a critical treatment step. Organic solids that remain in the system consume oxygen, contribute nitrogenous waste as they decompose and encourage heterotrophic bacterial growth. Microscreen drum filtration is widely used specifically to intercept these particles before they remain circulating within the system.

Once mechanically cleaned, the water can pass through a moving bed, a shower, or preferably both when the biological load justifies it.

The key is not simply installing all three filters. The key is designing the flow distribution, biological capacity, oxygen supply and maintenance strategy as one complete system.

Why the Three Filters Work So Well Together

Each filter addresses a different part of the waste stream.

There is some overlap.

A Bakki shower can perform considerable biological filtration. A moving bed can also help break down small amounts of organic material. An RDF may indirectly reduce ammonia production by physically removing organic waste before decomposition occurs.

But they should still be thought of as specialist components rather than three versions of the same filter.

The advantage of combining them is therefore not simply “more filtration.”

It is division of labour.

Start with the RDF: Remove Waste Before Trying to Process It

Consider a piece of uneaten koi food or a particle of faecal waste entering the bottom drain.

There are essentially two choices.

The filtration system can capture and remove that material quickly, or it can allow microorganisms to decompose it inside the system.

If it decomposes, oxygen is consumed and nutrients are released back into the water.

This is why good koi filtration attempts to physically export waste first rather than asking the biological filters to digest everything.

An RDF continuously passes pond water through a fine screen. As debris accumulates, water flow through the screen becomes restricted. The machine detects the increasing water-level differential and initiates a cleaning cycle, rotating the drum while high-pressure spray nozzles wash the accumulated waste into a separate waste channel.

Modern research into RAS drum filtration continues to demonstrate the importance of this mechanical stage. A 2025 study comparing different drum-filter screen sizes reported substantially lower particulate organic matter, particulate biological oxygen demand and particulate chemical oxygen demand in systems using drum filtration than in systems without it.

For the koi keeper, the practical consequence is important:

Every gram of waste physically removed from the RDF is waste that does not have to decompose somewhere else in the filtration system.

That creates a cleaner environment for the biological filters downstream.

Why Biological Filters Prefer Mechanically Clean Water

Nitrifying microorganisms are slow-growing compared with many microorganisms that consume organic carbon.

If a biological filter receives large quantities of faeces, uneaten food and fine organic debris, the surfaces intended for nitrifying biofilm can become increasingly occupied by faster-growing heterotrophic microorganisms.

Organic loading also increases biological oxygen demand.

Research on aquaculture biofilters shows that nitrification performance is influenced not only by ammonia concentration but also by dissolved oxygen, organic matter, temperature, pH, alkalinity, water movement and several other environmental conditions.

Mechanical filtration therefore protects biological capacity.

This relationship is one of the major reasons why an RDF followed by dedicated biological filtration has become such an effective architecture for intensive koi ponds.

The Moving Bed: A Stable Biological Engine

A moving-bed biofilter contains buoyant plastic media that is kept in constant motion, normally by vigorous aeration.

The media provides enormous collective surface area on which a biofilm develops.

Within that biofilm, ammonia-oxidising and nitrite-oxidising microorganisms participate in the nitrification process:

Ammonia → Nitrite → Nitrate

Ammonia is continuously produced by koi, particularly through their gills, and also enters the nitrogen cycle through decomposition of organic waste.

Without biological filtration, ammonia can accumulate.

The moving bed provides a controlled environment where nitrifying organisms can colonise the media rather than relying only on pond walls, pipework and other surfaces.

Moving-bed biofilm reactors are well established in recirculating aquaculture treatment systems, and reviews of MBBR technology highlight their effectiveness for nitrification and ammonia control.

Why the Media Must Actually Move

Adding moving-bed media to a chamber does not automatically create a good moving bed.

The chamber needs enough air to keep the media circulating continuously.

Ideally, the media should tumble throughout the chamber instead of forming stationary masses or dead zones.

The movement performs several functions.

It brings oxygenated water repeatedly into contact with the biofilm.

It distributes dissolved ammonia throughout the chamber.

It also causes the media pieces to contact one another, helping to control excessive biofilm thickness and exposing productive biofilm surfaces.

If too much media is packed into too small a chamber, the mass may no longer move correctly.

A smaller amount of properly fluidised media can perform better than a larger amount that barely moves.

Do Not Size a Moving Bed from Pond Volume Alone

This is one of the most important filtration-design concepts in this entire series.

Biological filtration demand is ultimately driven much more strongly by feeding and fish metabolism than by water volume alone.

Imagine two identical 30,000-litre ponds.

Pond A contains five koi and receives 200 grams of food each day.

Pond B contains forty rapidly growing koi and receives 1.5 kilograms of food each day.

The ponds have exactly the same water volume.

Their biological filtration requirements are obviously not the same.

The second pond produces much more nitrogenous waste and therefore requires substantially greater biological capacity.

For serious filtration design, consider: pond volume for hydraulic turnover; maximum realistic stocking level; maximum daily feed input; feed protein content; expected ammonia loading; operating temperature; desired ammonia and nitrite concentrations; and the tested performance of the chosen biological media.

Marketing figures quoting enormous theoretical surface area should not be treated as the only sizing metric.

What matters is how much biologically active protected surface is available and how the filter actually performs under the water-quality and loading conditions of the pond.

Nitrification Has an Oxygen Cost

A biological filter does not make ammonia disappear without consuming resources.

Nitrification requires oxygen.

The theoretical oxygen requirement for complete nitrification is approximately 4.57 grams of oxygen for each gram of ammonia-nitrogen oxidised to nitrate.

That oxygen requirement exists in addition to oxygen consumed by the koi and by microorganisms decomposing organic matter.

This is one reason heavily stocked ponds require substantially more aeration than lightly stocked ornamental ponds.

It also explains why biological filtration should never be designed independently from the pond's oxygen-management system.

Nitrification Also Consumes Alkalinity

Another frequently overlooked consequence of nitrification is loss of alkalinity.

The oxidation of ammonia ultimately generates acidity. Approximately 7.14 grams of calcium-carbonate-equivalent alkalinity are consumed for each gram of ammonia-nitrogen nitrified.

As feeding increases, nitrification increases.

As nitrification increases, alkalinity consumption increases.

If alkalinity becomes depleted, pH stability can deteriorate. Falling pH can then reduce biofilter performance and eventually create a dangerous situation for both the nitrifying biofilm and the koi.

This is why heavily stocked ponds should not merely monitor ammonia and nitrite.

KH or alkalinity and pH stability are also part of filtration management.

The Bakki Shower: More Than Another Box of Biological Media

A Bakki-style shower operates differently from a submerged moving bed.

Water is pumped to the top of the shower and distributed across media arranged in multiple trays or tiers. Water then falls through the media under gravity before returning to the pond.

The media therefore experiences repeated exposure to both water and atmospheric air.

From an engineering perspective, this is related to trickling biofiltration.

Trickling filters have a long history in recirculating aquaculture, and their nitrification performance is influenced by factors including dissolved oxygen, TAN concentration, organic loading, alkalinity, pH and hydraulic loading.

The exposed nature of a shower means oxygen is readily available to the biofilm when water distribution and ventilation are good.

At the same time, the falling water produces substantial air–water interaction.

Why a Shower Should Not Be Reduced to “Extra Oxygen”

A properly designed shower can contribute meaningful biological filtration.

But it should not be described simply as an oxygen generator.

Oxygen transfer depends on numerous factors including water temperature, current oxygen saturation, drop height, droplet formation, water distribution, air movement and hydraulic loading.

Water already close to oxygen saturation cannot simply absorb unlimited additional oxygen from atmospheric air.

The real strength of a shower is that it creates an environment with abundant air–water contact while supporting an aerobic biofilm.

The cascading water also assists gas exchange.

In high-performance koi systems, this combination makes showers particularly valuable alongside submerged biological filtration.

Moving Bed Versus Bakki Shower

Some koi keepers ask which one they should use.

For a demanding system, the more useful question is: Why choose only one if each offers different operational advantages?

A moving bed offers a large submerged biofilm protected inside continuously moving carriers. It can operate in a relatively compact chamber and does not require the same vertical pumping height as a large shower.

A shower provides an extremely oxygen-rich exposed environment but usually requires pumping water upward, which can substantially increase electrical consumption.

Using both can provide biological diversity and operational flexibility.

The important point is that adding a shower does not automatically compensate for an undersized moving bed, and adding a huge moving bed does not eliminate the need to consider oxygenation and gas exchange.

They are complementary tools.

Should Water Flow RDF → Moving Bed → Bakki Shower?

It can.

A perfectly functional system may operate: Bottom drains and skimmers → RDF → moving bed → pump → shower → pond.

This architecture is easy to understand.

Waste is removed first.

Mechanically clean water enters the moving bed.

The water then travels through the shower before returning to the pond.

However, there is a potential design problem.

The ideal flow through the RDF may not be the ideal flow through the moving bed, and the ideal moving-bed flow may not equal the desired shower flow.

Trying to force the entire pond circulation through all three components can therefore make one component dictate the operation of the others.

This is why a parallel arrangement deserves serious consideration.

A Better Architecture for Many High-Performance Ponds

A highly practical layout is: Pond → RDF → clean-water section.

After the RDF, the mechanically clean water is divided into separate treatment branches.

One branch feeds the moving bed.

Another branch feeds the Bakki shower.

Both branches eventually return to the pond.

Conceptually:

Bottom drains / skimmers ↓ RDF ↓ Clean-water chamber ↙ ↘ Moving bed Bakki shower ↓ ↓ Pond returns

This architecture allows the flow through each biological stage to be adjusted independently.

The moving bed can receive enough flow to continually deliver ammonia and oxygen without becoming hydraulically violent.

The shower can receive the higher flow required to wet its media effectively and produce strong air–water interaction.

Meanwhile the RDF can process the total mechanical circulation that the pond requires.

The Critical Principle: The RDF Should Still See the Dirty Water

Splitting flow is useful after mechanical filtration.

Splitting dirty pond water before the RDF requires greater caution.

If half of the circulation bypasses mechanical filtration and goes directly into biological chambers, much of the advantage of installing an RDF has been lost.

Solid waste will accumulate in the biological stage instead.

There are specialised aquaculture designs in which concentrated bottom-drain flows and cleaner side-stream flows receive different treatment, but these are engineered systems with deliberate solids-management strategies.

For conventional koi ponds, the simpler principle is generally safer:

Send the dirty pond water through effective mechanical separation before sending it to the biological filtration.

Flow Rate and Turnover Are Not the Same as Biological Capacity

A pond may circulate its entire volume once every hour.

That does not mean the biofilter is automatically large enough.

Similarly, a gigantic biological chamber will not compensate for poor pond circulation.

There are several separate design questions.

Pond turnover asks how frequently the pond water is circulated through the filtration system.

RDF flow capacity asks whether the drum can physically process that circulation without excessive water-level differential or constant cleaning.

Moving-bed capacity asks whether the biofilter can oxidise the ammonia load.

Shower flow asks whether enough water is distributed across the shower media to use it effectively.

These are related, but they are not interchangeable.

A good filtration system satisfies all four.

Do Not Chase Extreme Turnover Without a Reason

It is easy to assume that twice as much flow automatically means twice as good filtration.

That is not necessarily true.

Every increase in flow creates hydraulic consequences: larger pipes, greater friction losses, more pumping energy, higher RDF loading, greater return velocity, and potentially shorter residence time in some treatment stages.

The objective is not to move the maximum amount of water that the equipment can physically tolerate.

The objective is to move enough water to maintain water quality and collect waste efficiently without wasting energy or compromising treatment performance.

Why Pipework Matters as Much as Filter Size

A high-capacity RDF connected to undersized bottom-drain pipework will never perform to its theoretical capacity.

A powerful pump attempting to pull more water than gravity lines can provide may cause the clean side of a gravity-fed RDF to drop excessively.

Likewise, poorly designed pond returns can leave areas of the pond with weak circulation where solids settle instead of travelling toward the drains.

Filtration performance therefore begins inside the pond itself.

The system should create deliberate circulation patterns that encourage suspended and settleable waste to reach bottom drains and skimmers.

Only then can the RDF remove it.

The best mechanical filter in the world cannot capture waste that never reaches it.

Gravity-Fed RDF Systems

For larger purpose-built koi ponds, gravity-fed RDF configurations are particularly attractive.

The pond and dirty side of the RDF normally operate at approximately the same static water level.

Water leaving the clean side is pumped back to the pond.

As the RDF screen accumulates debris, resistance increases and the clean-side water level changes, triggering the cleaning cycle.

This arrangement allows bottom drains to feed the mechanical filter directly without chopping large waste particles through a pump first.

It also makes the RDF the natural first stage in a gravity-fed filtration architecture.

Where the Moving Bed Should Sit

After the RDF, a moving bed can be incorporated into a biological chamber within the gravity-fed filter system.

One common configuration is: Pond → RDF → moving-bed chamber → pump → pond.

If a shower is added, some or all of the pumped return can instead be sent upward to the shower.

Another arrangement uses independent pumps: RDF clean chamber → low-head pond returns; and RDF clean chamber → dedicated shower pump → Bakki shower.

This is particularly useful where the shower requires considerably more pumping head than submerged returns.

Rather than forcing the entire circulation through a high-head pump, only the shower circuit pays that energy penalty.

For ponds that operate continuously for many years, this distinction can have a large effect on electricity consumption.

Hydraulic Head Is an Operating Cost

A Bakki shower must normally sit above pond level because its final return depends on gravity.

Water therefore has to be lifted.

A pump moving 20,000 litres per hour against very little head can consume significantly less energy than a pump moving the same water to the top of a tall shower after accounting for vertical lift and pipe losses.

This does not mean showers are inefficient or undesirable.

It means the hydraulic system should be designed intelligently.

Oversized pipework, smooth bends, short runs where practical and properly selected pumps can substantially reduce unnecessary resistance.

Do not choose the pump only by the flow number printed on the box.

Pump performance changes with head.

The relevant figure is the actual flow produced at the total dynamic head of the installed system.

How Much Water Should Go Through the Shower?

There is no universal flow number that is correct for every shower.

The required flow depends on: shower dimensions; media type; media volume; tray design; number of tiers; water-distribution system; available ventilation; and the biological load.

The important objective is uniform wetting.

A pump may produce an impressive flow figure, but if most of the water travels down two narrow paths through the media, much of the shower is being underutilised.

This is called channeling.

Good distribution across the upper tray is therefore essential.

The shower should behave like a large active treatment surface, not several waterfalls cutting through isolated portions of dry media.

Avoiding Channeling in the Bakki Shower

Water naturally follows the path of least resistance.

If the top distribution pipe dumps large quantities of water into only a few areas, those areas become dominant flow channels.

Other media surfaces remain relatively dry.

This reduces effective contact area.

Good shower design therefore begins before the water touches the media.

The inlet should distribute water across as much of the tray width and length as reasonably possible.

The media arrangement should avoid creating tightly packed areas where water backs up while adjacent areas receive little flow.

The trays must also have enough drainage capacity that water does not simply accumulate and convert the shower into a partially flooded submerged filter.

Do Not Pack Shower Media Too Tightly

A shower requires both water and air.

Filling every available space with densely packed media may appear to increase surface area, but it can reduce airflow and create undesirable hydraulic conditions.

The ideal media arrangement should permit water to spread while allowing air to move through the structure.

Media selection should therefore consider more than advertised surface area.

Useful characteristics include: durability; open pore structure or protected surface; hydraulic openness; resistance to clogging; and suitability for biological colonisation.

There is no biological advantage in owning enormous theoretical surface area if much of it receives poor water flow or insufficient oxygen.

A High-Performance System Should Be Designed Around Feeding Rate

For serious koi keeping, filtration should be sized for the pond at its future peak load, not merely today's fish population.

Koi grow.

Collections expand.

Feeding increases.

A system that appears oversized when ten small koi are introduced may become marginal several years later when those same fish are much larger.

When planning FL17-type filtration, ask: What is the maximum quantity of food this pond may realistically receive on a heavy summer feeding day?

That number is far more useful than asking only: How many litres is the pond?

The pond volume determines hydraulic requirements.

The feeding load is a major driver of biological waste production.

Both must be considered.

Why Heavily Stocked Dealer Ponds Need More Safety Margin

Dealer and holding ponds are particularly demanding.

Fish numbers change rapidly.

New shipments may arrive.

Feeding can increase quickly.

Sedation, transportation and handling may add stress.

A filter sized to operate at its absolute theoretical maximum leaves little capacity for unexpected load.

Commercial and heavily stocked systems therefore benefit from excess biological capacity, additional aeration and multiple circulation paths.

The purpose of that margin is not simply to produce lower ammonia readings under normal operation.

It is to provide resilience when conditions are no longer normal.

Malaysia and Other Warm-Climate Considerations

Warm tropical climates deserve special attention because water temperature affects both the fish and the filtration system.

Warm water physically holds less dissolved oxygen than cooler water, while fish metabolism and biological oxygen demand can simultaneously increase. Aquaculture guidance consistently highlights the greater risk of oxygen depletion under warm conditions.

This makes generous aeration particularly valuable in heavily stocked Malaysian koi ponds.

The moving bed should receive reliable air continuously.

The pond itself should also have dedicated aeration rather than depending entirely on the filter.

The shower provides useful air–water contact, but it should not become the pond's only oxygen safeguard.

A blower failure, pump failure or blocked shower inlet should not be capable of removing virtually all aeration from the koi simultaneously.

Redundancy Is Part of Filtration Design

A sophisticated filtration system should be designed with failure in mind.

Ask what happens if the RDF stops rotating.

What happens if the RDF screen blocks?

What happens if the moving-bed air blower stops?

What happens if the shower pump trips?

What happens during a power failure?

What happens if one return line becomes blocked?

The objective is not to make failure impossible.

That is unrealistic.

The objective is to prevent a single failure from immediately becoming a pond catastrophe.

For valuable koi collections, redundancy may include independent air pumps, multiple water pumps, RDF overflow protection, bypass routes, emergency aeration and backup electrical power.

A filtration system's true quality becomes obvious not only when everything works, but also when something fails.

Do Not Make the RDF a Single Point of Catastrophic Failure

Because the RDF sits at the beginning of the system, a poorly planned failure mode can interrupt almost the entire circulation system.

The installation should account for possible screen blockage, sensor failure, drive failure, spray-pump problems and waste-tray blockage.

Many RDF systems incorporate overflow or emergency bypass arrangements.

The exact method depends on the equipment design, but the principle is essential:

Fish survival should not depend on one sensor and one motor working perfectly every minute of the year.

Regularly test the emergency behaviour rather than assuming it works.

Cleaning Frequency Is a Diagnostic Tool

An RDF that suddenly begins washing dramatically more frequently is telling you something.

Possible explanations include increased feeding, algae, a spawning event, disturbed sediment, increased fish load, a partially blocked screen or reduced clean-side water level.

The wash frequency is therefore more than a maintenance inconvenience.

It is information about what is happening within the pond.

Some modern installations monitor wash frequency and water level for this reason.

If the RDF begins cleaning continuously, investigate the cause instead of merely increasing the spray pressure and ignoring the underlying problem.

The Biological Filters Should Not Be Sterilised

One mistake sometimes occurs when UV, ozone or aggressive chemical treatments are added to an advanced filtration system.

The entire system is treated as though every microorganism is undesirable.

That is incorrect.

The moving bed and shower depend on microbial communities.

Medication, disinfectants, oxidising treatments and aggressive cleaning can potentially disturb biological filtration.

Whenever the pond requires treatment, consider the effect on the biofilter as part of the treatment plan.

After any major intervention, monitor ammonia and nitrite rather than assuming the mature biofilter remains unchanged.

A New Moving Bed Is Not an Instant Biofilter

Fresh plastic media may provide plenty of potential surface area, but it does not arrive with a mature nitrifying community.

Biofilter establishment takes time.

The Southern Regional Aquaculture Center specifically describes biofilter start-up as the process of developing nitrifying bacteria capable of processing the nitrogenous waste produced by the cultured animals.

Therefore, adding another 200 litres of new moving-bed media today does not necessarily create 200 litres of mature filtration capacity tomorrow.

The same applies to new shower media.

Biological capacity develops as the biofilm establishes.

During commissioning or major expansion, feeding should therefore be matched to the actual maturity of the filter.

Never Judge Biological Capacity by Water Clarity

Crystal-clear water can still contain ammonia.

This is one of the most dangerous misunderstandings in pond filtration.

The RDF can make the water visually beautiful while the biological filter is struggling.

Mechanical filtration and biological filtration solve different problems.

A serious pond should therefore be monitored using actual water tests.

At minimum, ammonia, nitrite, pH and KH or alkalinity should be understood and checked routinely, with nitrate, temperature and dissolved oxygen monitored according to the demands of the system.

Visual clarity is valuable.

It is not a substitute for water chemistry.

What Happens to Nitrate?

The combined RDF, moving bed and shower system is very effective at converting ammonia through nitrification.

But conventional aerobic biofilters primarily convert nitrogen rather than completely removing it.

Much of the final nitrogen remains in the system as nitrate.

Therefore, even an extremely sophisticated RDF–MBBR–shower system does not eliminate the need for broader water management.

Depending on the pond design this may include water changes, controlled continuous replacement, plant-based nutrient export or specially designed denitrification systems.

The correct method depends on the goals of the pond.

The important lesson is: zero ammonia does not mean zero nitrogen accumulation.

Example of a Balanced Integrated Layout

Consider a heavily stocked koi pond using bottom drains and a surface skimmer.

The bottom drains and skimmer feed an RDF by gravity.

The RDF removes the majority of easily captured suspended waste.

Clean water then enters a large clean-water or biological section.

A portion of that water passes through an aerated moving-bed chamber before being pumped to underwater pond returns.

A second pump takes mechanically filtered water from the clean side and sends it to the top of a Bakki shower.

The shower returns by gravity to the opposite side of the pond.

The pond returns create rotational circulation that encourages waste toward the bottom drains.

Air diffusers provide dedicated pond aeration.

The moving-bed blower provides independent biofilter aeration.

If one circulation pump stops, the second circuit can continue moving at least part of the pond water.

This is not the only correct arrangement.

But it illustrates an important philosophy: Mechanical filtration, biological filtration, gas exchange, circulation and redundancy are being designed together rather than assembled separately.

When a Straight-Line System Makes Sense

Despite the advantages of splitting flow, a serial configuration remains perfectly reasonable in some ponds.

If the hydraulic requirements are compatible, the following can be simple and effective: RDF → Moving bed → Pump → Bakki shower → Pond.

Advantages include simplified pipework and assurance that all circulated water experiences both biological stages.

It may be particularly practical when one pump can provide the required shower flow and the moving-bed chamber is large enough to handle the hydraulic throughput comfortably.

The important point is not that parallel filtration is always superior.

It is that designers should not assume a straight line is mandatory.

The correct configuration depends on the actual equipment and hydraulic requirements.

When Parallel Biological Circuits Make More Sense

A split system becomes particularly attractive when: the desired RDF flow is much greater than the appropriate flow through one biological chamber; the shower requires a different pump or substantially greater head; energy efficiency is important; redundant circulation is desirable; or the pond needs several separate return zones.

For large koi ponds, this flexibility can make the overall system easier to balance.

Common Mistake: Installing Enormous Filters but Poor Returns

A pond can contain an RDF, several hundred litres of moving-bed media and an enormous shower yet still collect debris in corners.

This happens because filtration equipment cannot compensate for poor pond hydraulics.

Waste must reach the filtration system.

Bottom-drain positioning, floor contour, return direction, surface circulation and skimmer placement therefore matter enormously.

A high-performance filter room should be designed at the same time as the pond shell.

Trying to repair poor circulation after the pond has already been constructed is often much harder.

Common Mistake: Making the Shower Too Difficult to Service

Biofilters eventually require inspection.

Pipes need cleaning.

Distribution holes can block.

Media can shift.

Pump strainers require maintenance.

If accessing the shower requires dismantling half of the pond landscaping, maintenance will gradually be postponed.

The same principle applies to the RDF and moving bed.

Good engineering considers maintenance access from the beginning.

Filters that are easy to service are more likely to remain effective over many years.

Common Mistake: Depending on Manufacturer Maximum Flow Figures

Equipment specifications frequently quote maximum flow.

That value should not automatically become the design flow.

Maximum performance may assume ideal water levels, clean screens, specific pipe sizes and almost no installation losses.

Real ponds contain bends, valves, vertical lift and fluctuating water levels.

A sensible design therefore includes hydraulic margin.

Running every component continuously at its theoretical maximum leaves little room for fouling, future expansion or unexpected operating conditions.

Common Mistake: Adding Biological Media Instead of Fixing Mechanical Filtration

When ammonia or water quality deteriorates, some keepers immediately add more biological media.

Sometimes that is appropriate.

But if the existing biological chambers are full of trapped organic waste because mechanical filtration is inadequate, simply adding more media may not address the fundamental problem.

Inspect the entire treatment chain.

Waste capture. RDF operation. Water distribution. Aeration. Media movement. Feeding. Alkalinity. Flow.

Only after identifying the limiting factor should equipment be added.

Common Mistake: Believing More Media Is Always Better

A biological chamber has an optimum operating range.

Too little media provides insufficient surface area.

But too much can restrict movement.

Similarly, too much shower media can restrict drainage and airflow.

The goal is therefore not to maximise the physical amount of plastic or ceramic material inside the filter.

The goal is to maximise the amount of productive biological surface operating under good conditions.

Common Mistake: Forgetting Future Expansion

Pipe diameter, RDF capacity and chamber dimensions are difficult to change after construction.

Pumps and media quantities are comparatively easy to change.

Whenever practical, build permanent infrastructure with some allowance for expansion.

An oversized biological chamber can initially contain less media and receive more later.

A larger RDF can operate below its maximum capacity.

Large gravity lines can carry lower flow.

But an undersized concrete chamber or pipe buried beneath the pond is far more difficult to correct.

Designing for Efficiency Rather Than Equipment Count

A filter room containing ten different devices is not automatically better than one containing four.

Every additional device adds: resistance, electricity consumption, maintenance, failure points, and complexity.

The objective is not to build the most complicated filtration system possible.

It is to achieve excellent water quality using the simplest reliable architecture that provides the required treatment capacity.

RDF + moving bed + Bakki shower works particularly well because the three components already cover several major filtration requirements without excessive overlap.

How to Evaluate the Completed System

Once commissioned and biologically matured, judge the system using measurable outcomes.

Water should remain clear.

Solid waste should leave through the RDF rather than accumulating throughout the biofilter.

The moving-bed media should circulate properly.

The shower should receive even water distribution.

Ammonia and nitrite should remain under control at peak feeding.

pH and alkalinity should remain stable.

Dissolved oxygen should remain strong, particularly during warm weather and early morning.

RDF wash frequency should remain within a sensible operating range.

The filter room should be serviceable without major disruption.

And most importantly, the system should retain some ability to protect the fish when individual components fail.

That is a much better definition of a high-performance filtration system than simply counting how many filters have been installed.

Frequently Asked Questions

Do I need both a moving bed and a Bakki shower?

Not every pond does. Either can provide substantial biological filtration when properly designed. However, demanding or heavily stocked ponds can benefit from using both because they provide biological filtration under different operating conditions and create additional treatment capacity and resilience.

Should the RDF always come first?

In most conventional koi systems, effective mechanical filtration should occur before the main biological filtration so that solids are removed rapidly rather than decomposing inside the biofilter. Specialised split-flow aquaculture designs exist, but their solids-management systems are intentionally engineered.

Can the Bakki shower replace the moving bed completely?

It can provide significant biological filtration, but whether it can replace a moving bed depends on biological loading, media quantity, flow, temperature, oxygen conditions and operating goals. For heavily stocked systems, maintaining both can provide useful additional capacity and redundancy.

Can I run the whole pond through the Bakki shower?

Yes, if the hydraulic design, pump capacity and shower dimensions support the required flow. However, it is not mandatory. A separate shower circuit can often provide greater flexibility and lower overall pumping energy.

Does an RDF remove ammonia?

No. An RDF removes particulate matter. By removing organic waste early it can reduce the amount of material that later decomposes and contributes to nitrogenous loading, but dissolved ammonia requires biological or other suitable treatment.

Does a moving bed remove nitrate?

Conventional well-aerated moving beds primarily promote nitrification, converting ammonia through nitrite to nitrate. They should not normally be expected to provide major nitrate removal.

Does the Bakki shower remove nitrate?

A conventional highly oxygenated shower is primarily an aerobic treatment environment. It should not be treated as a reliable nitrate-removal system. Nitrate control normally requires separate nutrient-export strategies.

Should all water leaving the RDF pass through biological filtration?

The system should have enough total biological treatment capacity for the pond's ammonia load, but every litre does not necessarily need to pass through every biological device during every circulation. Parallel treatment circuits can operate successfully when properly sized.

Is a larger RDF always better?

Additional capacity can be useful, particularly for future expansion, but an RDF still needs suitable pipework, water levels and hydraulic design. Buying a huge drum filter does not solve undersized bottom drains or poor pond circulation.

What should be oversized first when building a new pond?

Permanent infrastructure deserves particular attention: gravity lines, filter chambers, drain capacity and accessible pipework are difficult to change later. Pumps, air pumps and biological media can usually be upgraded more easily.

Final Perspective

The RDF, moving bed and Bakki shower should not be viewed as three independent filters purchased one after another.

They should be viewed as stages within one engineered water-treatment system.

The RDF's job is to remove physical waste quickly.

The moving bed provides a stable submerged environment for nitrification.

The Bakki shower adds another highly aerobic biological environment together with intensive air–water contact.

But the real performance comes from the way the components are connected.

A poorly designed system can have expensive equipment but weak circulation, insufficient oxygen, excessive pumping costs and inadequate biological capacity.

A well-designed system balances:

waste removal → biological conversion → oxygenation and gas exchange → hydraulic circulation → redundancy

The most sophisticated filtration systems often appear simple because every component has a clearly defined job.

For a heavily stocked koi pond, one of the strongest configurations is therefore not necessarily:

RDF → Moving Bed → Bakki Shower in one compulsory line

but rather:

capture the solids with the RDF first, then distribute clean water intelligently through appropriately sized biological treatment circuits.

Once this principle is understood, RDFs, moving beds and Bakki showers become much easier to combine effectively.