CO2 & Lighting: CO2 Bubbles-Per-Second Starting Points Guide for Aquarium Keepers
Learn why CO2 bubbles per second is a flow proxy rather than a concentration target, compare diffuser vs reactor efficiencies, and dial in safe starting BPS for US, UK, and EU tanks.
Updated 2026-11-20 · 1697 words
Every new pressurized CO₂ user asks the same question on day one: "How many bubbles per second should I set for my tank?" While a bubble counter is indispensable for making repeatable needle-valve adjustments, bubbles per second (BPS) is a gas-flow proxy, not a concentration measurement. Two aquariums of identical volume running at 2.0 BPS can sit at 12 ppm CO₂ in one living room and 38 ppm in another depending on bubble-counter fluid viscosity, diffuser dissolution efficiency, canister flow rate, surface ripple, and plant biomass. Treating BPS as a conservative starting point—and tuning from there using a 4 dKH drop checker and degassed pH differential—keeps livestock safe while dialing in rapid plant pearling.
Why One Bubble Is Not a Universal Unit of Gas
A bubble counter simply forces CO₂ gas through an orifice submerged in liquid. The actual mass of carbon dioxide inside each rising bubble varies by up to 60% across different hardware:
- Orifice diameter and counter fluid: Narrow glass needle tips filled with pure water produce smaller, faster bubbles than wide plastic tubing ports filled with viscous mineral oil or glycerin. Glycerin slows bubble detachment and can yield larger individual gas volumes per count.
- Working pressure and backpressure: Standard in-tank ceramic disk diffusers require 15–25 psi (1.0–1.7 bar) just to overcome ceramic pore tension, whereas inline atomizers often require 30–40 psi (2.1–2.8 bar). Higher downstream backpressure compresses gas inside the counter chamber, packing more moles of CO₂ into each visible bubble.
- Temperature and gas density: At 22 °C and 1 atm, CO₂ gas has a density of roughly 1.83 mg/mL. A typical 3 mm diameter sphere has a volume of ~0.014 mL (~0.026 mg CO₂), whereas a 4 mm bubble holds ~0.034 mL (~0.062 mg CO₂)—more than double the mass for the exact same "1 BPS" count.
Dissolution Efficiency Across Diffuser and Reactor Types
How much of that injected gas actually dissolves before escaping at the water surface determines your required BPS rate:
- In-tank ceramic diffuser (45–65% efficiency): Produces a fine mist of microbubbles, but 35–55% of gas still coalesces and pops at the surface unless pushed downward by strong outflow circulation.
- Inline canister atomizer (75–85% efficiency): Injects microbubbles directly into pressurized return tubing, giving extra contact time inside the hose and creating a fine "Sprite-like" haze across the entire tank.
- External Cerges / Rex Grigg / Mazzei reactor (95–99% efficiency): Traps CO₂ gas inside a vertical chamber against downward canister flow until 100% of the gas dissolves into solution with zero visible mist. Because virtually no gas is wasted to the room, reactors often achieve 30 ppm CO₂ at 30–40% lower BPS than an in-tank ceramic disk.
Conservative BPS Starting Points by Tank Volume and Diffusion Method
Always begin on the conservative side during week one, especially when stocking sensitive fish (Otocinclus, discus, rummy-nose tetras) or Caridina shrimp. The starting baselines below assume moderate surface agitation (gentle ripple without splashing) and a 2-hour pre-light lead time.
| Net Water Volume | In-Tank Ceramic Diffuser | Inline Atomizer | External Reactor (95%+) | Primary Tuning Guardrail |
|---|---|---|---|---|
| 20–40 L (5–10 US gal) | 0.3–0.5 BPS (1 bubble / 2–3 s) | Rarely used | Rarely used | Use micro-needle valve; nano volumes swing fast |
| 60–90 L (15–24 US gal) | 0.8–1.5 BPS | 0.6–1.0 BPS | 0.5–0.8 BPS | Place ceramic disk directly under lily pipe outflow |
| 100–160 L (26–42 US gal) | 1.5–2.5 BPS | 1.0–1.8 BPS | 0.8–1.5 BPS | Watch surface film; avoid dead zones behind rockwork |
| 180–260 L (48–68 US gal) | 2.5–4.0 BPS | 1.8–3.0 BPS | 1.5–2.5 BPS | Count 10-second intervals (e.g., 25 bubbles / 10 s) |
| 300–450 L (80–120 US gal) | 4.0–6.0+ BPS | 3.0–4.5 BPS | 2.5–3.5 BPS | Inline atomizer or external reactor strongly preferred |
7-Day Incremental CO₂ Tuning Protocol
Never crank a needle valve up to 4 BPS on day one and walk away. Follow a disciplined multi-day ramp:
- Day 1 (Baseline): Set your conservative starting BPS using a stopwatch over 15 seconds (e.g., 15 bubbles in 15 seconds = 1.0 BPS). Turn the solenoid on 90 minutes before lights-on. Observe fish respiration at hours 2, 4, and 6.
- Day 2 (Saturation Check): Inspect the 4 dKH drop checker at midday and measure tank pH versus a degassed sample. If the drop checker remains deep blue and the pH drop is under 0.5 units, increase flow by +0.3 to +0.5 BPS (a tiny 1/16th turn of a quality needle valve).
- Days 3–5 (Stepwise Increases): Wait 24 hours between adjustments so working pressure and tank equilibration settle. Increase by +0.3 to +0.5 BPS per day until the drop checker turns crisp lime green and the degassed-to-saturated pH drop reaches 0.9–1.0 units.
- Day 6–7 (Flow Optimization Before More Gas): If you hit 3–4 BPS on a 120 L tank and the drop checker stays blue-green, stop adding gas and inspect circulation. Move the diffuser under the filter return so microbubbles sweep across the full tank length before rising.
Regional Worked Examples: US, UK, and EU Setups
1. United States Example: 29-Gallon Tall (110 L) With In-Tank Ceramic Diffuser
A keeper in Chicago runs a standard 29-gallon tall aquarium (30" × 12" × 18", net water 95 L / 25 US gal) lit at 60 µmol·m⁻²·s⁻¹ with a hang-on-back filter and a 25 mm ceramic diffuser.
- Starting Point: For 95 L net volume with an in-tank ceramic diffuser and moderate HOB surface splash, the table recommends 1.2–1.5 BPS. The keeper sets the needle valve to 12 bubbles per 10 seconds (1.2 BPS).
- Midday Audit: After 3 days, the drop checker sits at dark forest green and the pH drop from degassed Chicago tap water (pH 7.80) reaches only 0.65 units (pH 7.15), indicating ~13–15 ppm CO₂. Because the HOB waterfall strips gas aggressively, the keeper raises the water level to within 1 cm of the spillway to quiet the splash and nudges the needle valve to 1.8 BPS. Within 48 hours, pH drops to 6.82 (0.98-unit drop) and Rotala pearling begins cleanly.
2. United Kingdom Example: 180 L Community Tank With Inline Atomizer
An aquarist in Manchester runs a 180 L (net 155 L / 41 US gal) planted community tank filtered by an Oase BioMaster 350 canister running 16/22 mm hoses and an inline CO₂ atomizer at 32 psi working pressure.
- Starting Point: Because inline atomizers achieve ~80% dissolution efficiency, the keeper starts at 1.5 BPS (15 bubbles in 10 seconds) rather than the 2.5 BPS an in-tank disk would need.
- Fine-Tuning: At 1.5 BPS, the tank achieves a 0.78 pH drop (from degassed 7.52 to 6.74). Increasing to 2.0 BPS brings the tank to pH 6.53 (0.99-unit drop) with a bright lime-green drop checker and zero distress from cardinal tetras or Corydoras.
3. European Union Example: 300 L Dutch Aquascape With External Reactor
A hobbyist in Utrecht, Netherlands maintains a densely planted 300 L (120 × 50 × 50 cm, net 260 L / 68.7 US gal) Dutch-style tank powered by a dual-filter loop and a vertical Cerges external CO₂ reactor.
- Starting Point: At 98% dissolution efficiency, virtually zero gas escapes un-dissolved. The keeper begins at 2.2 BPS (counting 22 bubbles over 10 seconds) using a precision Swagelok-style metering valve.
- High-Biomass Adjustment: As dense stem bushes of Ludwigia, Bacopa, and Pogostemon triple in mass over six weeks, photosynthetic carbon demand rises. The keeper steps up gradually from 2.2 BPS to 3.2 BPS—a rate that still uses 35% less cylinder CO₂ per month than an in-tank ceramic diffuser would require for the same 32 ppm target.
Gas-Line Physics: Tubing, Needle Valves, and Bubble-Counter Fluids
Even before gas reaches your diffuser, the plumbing between your regulator and aquarium shapes stability and bubble size. Standard soft airline silicone tubing is highly permeable to carbon dioxide; over 2 meters of silicone line, up to 15–25% of low-pressure CO₂ can diffuse directly through the tubing walls into the room, causing overnight pressure bleed and sluggish morning startup. Always plumb CO₂ lines with rigid polyurethane (PU) or PTFE-lined 4×6 mm tubing rated for 60+ psi.
Inside the bubble counter, the choice of fill fluid changes both evaporation rate and bubble volume:
- RO or tap water: Low viscosity (~1.0 cP at 20 °C) yields crisp, small bubbles that detach rapidly, but water evaporates within 2–4 weeks and can grow white bacterial slime.
- Pure vegetable glycerin or mineral oil: High viscosity slows bubble rise, making individual bubbles easier to count visually above 3 BPS, and never evaporates. However, high viscosity produces slightly larger bubbles per count—meaning 2.0 BPS in glycerin can deliver 15–20% more CO₂ mass than 2.0 BPS in water.
Diagnosing Flow Dead Zones vs. Insufficient Injection
Before raising your needle valve past 3–4 BPS on a medium tank, perform a 30-second dye or microbubble flow trace. Watch the fine CO₂ mist leaving your ceramic diffuser: if the plume rises straight up to the water surface in a narrow column and pops, more than 60% of your gas is venting into the room. Positioning the diffuser beneath the downward sweep of a lily pipe or spray bar forces microbubbles horizontally across the entire tank length, extending bubble residence time from 1.5 seconds to 6–8 seconds and often raising dissolved CO₂ by 8–12 ppm at the exact same BPS setting.
Cylinder Sizing, Refill Intervals, and Mass Consumption Math
Understanding how BPS translates into monthly CO₂ consumption helps you size cylinders appropriately. At 2.0 BPS in water (~0.030 mg CO₂ per bubble), an 8.5-hour daily injection schedule delivers 2.0 × 3,600 × 8.5 × 0.030 mg ≈ 1,836 mg (1.84 g) of CO₂ per day, or roughly 55 g per month. A 5 lb (2.27 kg) US aluminum cylinder or a 2 kg European/UK steel bottle therefore lasts over 36 months on a 120 L tank with an efficient reactor, whereas a small 95 g disposable cartridge runs dry in just 6–7 weeks—making bulk refillable cylinders dramatically cheaper per gram of carbon delivered.
Diagnosing Regulator Creep and Needle-Valve Hysteresis
If your bubble count reads 1.5 BPS at 09:00 and climbs to 2.4 BPS by 14:00 without anyone touching the knob, suspect either temperature-driven diaphragm expansion in a single-stage regulator or debris lodged in the needle-valve seat. Shut off the main cylinder valve, bleed residual line pressure, unscrew the metering stem gently if serviceable, and blow out dust with compressed air. Never apply petroleum grease to CO₂ regulator threads; use only oxygen-safe PTFE tape on high-pressure fittings.
Bubble-Counter Safety & Regulator Maintenance Checklist
- Count over 10–15 seconds above 2 BPS: Human eyes miscount rapid streams above 2.5 BPS. Time 10 seconds on a phone stopwatch and divide by 10 for accurate repeatability.
- Install a CO₂-rated check valve: Place a brass or spring-loaded check valve between the bubble counter and the tank so siphon backflow never floods your solenoid or needle valve when the cylinder turns off.
- Inspect ceramic disks monthly: Biofilm and green spot algae clog ceramic pores, raising backpressure and shrinking effective flow. Soak dirty disks in 1:10 household bleach for 15 minutes, rinse thoroughly, and soak in double-dose dechlorinator before reinstalling.
- Watch for end-of-tank dump (EOTD): On single-stage regulators, dropping cylinder pressure below 500 psi (35 bar) can cause working pressure to spike, suddenly doubling your BPS. Refill cylinders as soon as the high-pressure gauge enters the red reserve zone.
- Verify flow stability 30 minutes after adjustment: Brass needle valves experience slight O-ring compression creep; always re-count bubbles 30 minutes after touching the knob.