-1.3 Carbon Dioxide Removal

1.3 Carbon Dioxide Removal

The carbon dioxide removal system, or scrubber, is the heart of a rebreather device. Proper use of the scrubber requires utmost understanding in how chemical absorption takes place, and that the design of a scrubber canister factors in to maximally utilizing the sorbent media. The current state of the art involves the use of pelletized calcium hydroxide which indirectly reacts with carbon dioxide via the use of water and a sodium hydroxide catalyst. This has been tried and true chemistry used in diving and medical applications for over a century.

 

 

Chemical reaction for carbon dioxide absorption within scrubber. From IANTD 2001.

 

Importantly, the scrubber reaction is exothermic (produces heat) and generates water as a byproduct. Breathing warm/moist gas can be an advantage to the diver who may likely be immersed in water cooler than his/her body temperature. This warm gas helps with heat retention, as opposed to the dissipative heat loss that comes with breathing cold and dry gas from a cylinder using open circuit diving. Scrubber heat can be preserved by insulating the canister within the breathing loop space itself (scrubber-in-lung, SIL) or potentially encasing the scrubber canister in an insulative wrap such as neoprene, cast foam, or similar. There is also evidence suggesting that warmer CaOH based scrubbers are more efficient at binding carbon dioxide, therefore offering improved performance characteristics.

A prototype scrubber canister incorporating a cast foam insulative barrier. This concept was abandoned in our RD1 development project given costs exceeded benefits.

 

A well-designed scrubber canister should accommodate accumulated water via a trap, absorbent pad, and/or evacuation mechanism in a position that does not impede oxygen sensor function on the inhalation side and does not impede gas flow through the scrubber. By absorbing or having the ability to evacuate this water byproduct at the exhale side of scrubber, the scrubber material can be kept reasonably dry and prevent caustic soda from being formed within the breathing loop.

In addition to water from the scrubber reaction, or ‘metabolic water’, water might also be introduced through accidental loss of the diver’s mouthpiece, or simply loose lips. This water would ingress on the exhale side of the scrubber, emphasizing the benefits of water trap placement on the exhalation side of the breathing loop. The scrubber bed itself should be additionally protected. This may be possible with a strategically placed baffle, a volume of space sufficient to accommodate some nominal water intrusion, perhaps a hydrophobic membrane, or simply inlet and outlet loop connections that are not blocked by intruded water.

Generally speaking, a scrubber is just a can. The above features aid in accommodating the realities of water ingress and benefitting from temperature. The size and shape of the can are where things get more interesting. An essential consideration is that the full scrubber is designed such that gas flow resistance is kept to a minimum. Sorbent retention screens are used at the inlet and outlet and should be of a mesh size that doesn't impede gas flow. In the case of our RD1 development program, the cross-sectional area of the sorbent bed, while accounting for intergranular space, is roughly equal to the breathing loop nominal diameter of 1.5". Absent intruded water, the scrubber itself does not add much to the resistive work of breathing through the loop. Many scrubbers on the market are too small in diameter - their resistive work of breathing should be measured and well considered as it applies to CO2 retention during more arduous work.

Rendering of a concept canister for the RD1 development project, which incorporated a spring tensioned scrubber bed, rather than threaded rod compression.

Scrubber Performance

Running out of air (open circuit) is the worst feeling in the world and, of course, best avoided with proper planning. It is equally devastating to run out of scrubbing capacity (closed circuit). Unfortunately, the diver wouldn’t be afforded much time to frantically find a buddy, as he or she would be inspiring quickly elevating CO2 and end up blacked out on the bottom before realizing what happened. Logic again, tells me to make certain that this doesn’t happen through proper planning, though also through a deeper appreciation for scrubber performance limitations.

In the absence of true scrubber chemistry monitoring, which would be a complex realization at best, we are left with educated estimates requiring a degree of conservatism in dive planning. Rebreather divers are broadly taught that scrubber absorbent is worth ‘one hour per pound’. This is generally accurate, but what about the effects of depth? What about the metabolic rate for the individual diver and varying metabolism through different work rates? What about the behavior of axial versus radial designs? What if some water is introduced? What about channeling through the bed? All of these variables, and more, have an impact on scrubber duration.

Scrubber Capacity

Given the above stated variables contributing to scrubber durations, the maximum allowable scrubber duration or endurance of the rebreather should be engineered such that this ungaugable filter is treated with conservatism, and oxygen available becomes the limiting consumable. By keying the oxygen supply volume carried with the diver’s metabolism to a buffered theoretical scrubber capacity, critical conservatism is embraced. Frequently rebreathers carry too much oxygen which can be cause for bad habits in over-extending the scrubber capacity - the premise being to permit multiple dives on one cylinder. Well great - do you have enough scrubber capacity to match? It's unlikely.

Practically speaking, a small 2-liter cylinder is more than adequate for even long duration surface-to-surface excursions, and when calculated out, provides for about 5 hours of dive time (300 minutes) based on a metabolic consumption rate of 1 liter per minute. So, given the rule of thumb for scrubbers (one hour per pound) we’ve used in practice for decades, a unit with a 2-liter oxygen cylinder should incorporate a scrubber bed that contains a minimum of five pounds of absorbent. This represents a properly balanced system which forces the diver to replenish both oxygen and scrubber consumables together. Many units promote this logic through their engineering, though some do not.

Temp stick technology has allowed divers to push this five-hour mark and provides insights in to how variable metabolic consumption, depth, and dwell time (huffing and puffing versus relaxed breathing) factors into scrubber duration. While we assume a diver consumes 1 liter per minute of oxygen, and that is the threshold commonly used when learning about constant mass flow orifices in mCCR training and diving, a very relaxed human might only metabolize 0.25 liters per minute. That would vastly expand the oxygen life support capacity afforded by the little 2-liter cylinder, and so we have to be careful matching the scrubber capacity accordingly and not making assumptions in the wrong direction.

absorbent weight/volume					axial scrubber diameter
sofnolime is 1g/cm3 granular state					10.16 cm	15.24 cm	20.32 cm	25.40 cm
lbs	kg	g	cm3		4" diameter	6" diameter	8" diameter	10" diameter
3.3	1.5	1500	1500	Bed length    cm	18.5	8.2	4.6	3.0
4.4	2	2000	2000		24.7	11.0	6.2	3.9
5.5	2.5	2500	2500		30.8	13.7	7.7	4.9
6.6	3	3000	3000		37.0	16.4	9.3	5.9
7.7	3.5	3500	3500		43.2	19.2	10.8	6.9
8.8	4	4000	4000		49.3	21.9	12.3	7.9
9.9	4.5	4500	4500		55.5	24.7	13.9	8.9
11.0	5	5000	5000		61.7	27.4	15.4	9.9
12.1	5.5	5500	5500		67.8	30.2	17.0	10.9
13.2	6	6000	6000		74.0	32.9	18.5	11.8
14.3	6.5	6500	6500		80.2	35.6	20.0	12.8
15.4	7	7000	7000		86.3	38.4	21.6	13.8
16.5	7.5	7500	7500		92.5	41.1	23.1	14.8
17.6	8	8000	8000		98.7	43.9	24.7	15.8
18.7	8.5	8500	8500		104.8	46.6	26.2	16.8
19.8	9	9000	9000		111.0	49.3	27.8	17.8
20.9	9.5	9500	9500		117.2	52.1	29.3	18.8
22.1	10	10000	10000		123.3	54.8	30.8	19.7

.

Scrubber weight, fluid volume, and axial design dimensions. Yellow indicates scrubber size required to accommodate a maximal tidal volume of 2.5L. Green indicates scrubber size required to accommodate an average, though till conservative tidal volume.

 

For reasons including gas velocity through the scrubber bed, intergranular space considerations, and dwell time, I am an advocate for a scrubber containing a volume of sorb in between the average tidal volume of an adult male (0.5 to 1 fluid liter) and the average maximum vital capacity of an adult male (about 4 fluid liters). Given the intergranular space within carbon dioxide absorbent being roughly 50% of its bulk volume, this delta, of say 2.5 fluid liters which represents a volume of breath exchange for a human hard at work, would require a 5-fluid liter scrubber bed to allow for adequate gas dwell time under all working conditions while not permitting premature breakthrough during hard work. This is a scrubber of about 11 pounds.

Yes, that’s huge, and no, it's not practical.

However, now consider the multitude of variables referenced, and that this monster accounts for the extreme end of work that the scrubber may be forced to do. It would be very difficult to impossible for the average human to sustain 2.5 liters per breath for any length of time. Huffing and puffing like this is ill-advised, and if encountered, is certainly good reason to go off the loop to relax. Good form in rebreather diving is slow and methodical, not aggressive huffing and puffing.

Rebreather diving is a finesse activity, like a ballet – not a mosh pit.

For consideration, 3^rd party rebreather testing often takes place with a breathing cycle of 40 liters per minute. With an average respiratory rhythm of say 15-20 ventilations per minute, that is 2-2.67-liter ventilations. That is consistent with our 2.5 liters per breath estimate, and with being able to account for this volume within the intergranular space of the absorbent, there is virtually zero risk of premature breakthrough at the 5-hour mark (when oxygen would be depleted), and it provides for substantial buffer in the event of a too deep/too long contingency. When considering the future of extended range diving – where the enclosed design principles are targeted to take us, this scrubber capacity just makes me feel that much better.

Now however in practical terms, this monster 11-pound scrubber would never be utilized. More practically, a scrubber of about half this size is the standard – about 5-5.5 pounds. This size more appropriately accounts respiratory cycles consistent with our average tidal exchanges, with limited peaks from arduous work or exercise during the earlier parts of a dive (which is more common and consistent when considering much shallow time is often spent leisurely decompressing). Exhale counterlungs help to reduce velocity (increasing dwell time) and are a must have on mixed-gas/deeper diving units in particular.

Understanding Theoretical Bed Life (TBL)

As of this writing, there are no widely accepted means for effectively gauging what is actually going on within a scrubber. Is this a problem, or not? If you ask me, it's a scary thought - scary enough to give me great pause in diving rebreathers at all, which has been the case from time to time.

 

Three known options do exist for determining scrubber use in both atmospheric and hyperbaric systems. These include (Wallace, 2005):

 

1. Use of a clear canister and indicator chemical.  Chemical changes color as absorbent becomes saturated with CO2.
2. Use of a temperature sensor on or in canister.  Sorb releases heat as CO2 is        captured and cools down when chemical is exhausted, revealing a reaction front tracking through the bed over time.
3. Calculate dive time using absorbent specifications and design parameters of your scrubber canister (volume, channeling, efficiency), and key this to metabolic production of CO2 (consumption of O2).

Some efforts have been made to design CCR scrubbers with the above in mind, particularly to measure use of the scrubber media itself using its exothermic reaction to monitor a reaction front across the scrubber plenum. This is interesting to observe, particularly how the reaction front moves forward at depth, and then regresses on ascent; reaffirming that ‘bulk loading’ is a real thing (to be discussed).

Carbon dioxide can bypass the scrubber media in one of several ways. These include physical channels or voids in the scrubber media, inadequate dwell time of exhaled gas within the scrubber bed and simply overextending the remaining bed life by speed and volume – all rendering the scrubber useless. These variables cannot be pre-determined or used in calculations as each independent canister fill may provide unique inadequacies.

In practice, divers load a scrubber canister with absorbent material (generally granular calcium hydroxide) and dive to the recommendations of the manufacturer, which may or may not have been actually verified analytically in a laboratory. During a real dive event, the variables affecting a scrubber’s performance are hardly absolute and vary widely. Again, variables affecting scrubber performance include divers’ breathing rate changes, surrounding water temperature, insulative capacity of the scrubber canister, bed length of absorbent material, velocity of gas passed through the scrubber, and gas density changes with depth, among others. 

Given these factors, simply stating that a scrubber is “rated for 4 hours regardless of depth” is a grossly inadequate misrepresentation of the scrubber’s capabilities and undoubtedly a factor in irresponsibly pushing scrubber endurance – surely a contributor to past incidents. The best situation is that actual laboratory and in-water tests are performed by rebreather manufacturers to determine the Theoretical Bed Life (TBL). TBL is a measure of the absorption capacity of a given amount of scrubber media under a specified workload while considering fixed environmental conditions (Nuckols et al., 1996). 

In Europe, CE Standards are in place to provide performance benchmarks through testing and provide a baseline standard for which to infer that a scrubber is capable of operating. Manufacturers are required to meet this CE standard to sell to European markets. In the US, there is no required testing by rebreather manufacturers; however, the National Oceanographic and Atmospheric Administration (NOAA) has proposed a set of testing criteria (2004), similar to CE, to guide manufacturers in developing rebreathers acceptable for use by NOAA and other federal agencies. According to NOAA, rebreather scrubbers should be evaluated and ‘rated’ while subjected to the following conditions:

1. CO2 injection:  1.35 standard liters per minute (SLPM)
2. ventilation rate of 40 l/min. The diluent gas will be air or nitrox
3. water temperature: 40 +/- 1 F (3.9-5.0 C)
4. depths of 60, 100, and 165 fsw

The two listed human factors of CO2 injection volume (simulated as post-metabolic exhalation during tests) and ventilation rates are, by design, very conservative variables. In particular, CO2 injection of 1.35 SLPM is approximately fifty percent more than the VCO2 of a diver under normal conditions (0.8 to 1.0 SLPM - the average we’ve used throughout this text as our noted 5-hour operating benchmark assuming a 1:1 oxygen metabolic consumption rate to VCO2 production).

Properly rating a scrubber should reflect the TBL, under specific controlled conditions, which include a specified depth. While this is most accurately determined analytically in a laboratory and is specific to the rebreather in its entirety due to gas or fluid dynamics, it is possible to calculate TBL for a specified volume, or mass, of absorbent material. It must be considered, however, that this calculation does not consider canister geometry, flow paths, or other variations from a canister design perspective.

Here we propose that by working through a series of calculations described in Nuckols et al. (1996), TBL can be determined for any given volume of absorbent, and with a degree of conservatism by utilizing the testing variables proposed by NOAA (2004) within the calculations. The utility value in these calculations includes being able to gain a reference point for dive planning for any given scrubber volume, across a spectrum of dive depths.

The Nuckols Equations

Nuckols et al. (2006) presents a set of calculations to determine Theoretical Bed Life (TBL). Theoretical bed life (in hours) is equal to the theoretical absorption capacity divided by the metabolic load. Theoretical absorption is determined according to Equation 1.

Klos et al. (2004) describes the reaction between soda lime and carbon dioxide in Equation 2 and goes on to present the sorbent reaction in Equation 3.

The first equation demonstrates the reaction of the Carbon Dioxide with Sodium Hydroxide. The main component of the soda lime is Calcium Hydroxide, which is shown in the second equation. From this, one can see the 1:1 molar ratio between the absorbent (Ca(OH)₂) and the carbon dioxide (CO₂) which was used in our calculations. Nuckols et al. (1996) describes metabolic load, m_co2, in Equation 4 where V₀₂ is the oxygen consumption rate, RQ is the respiratory quotient, and ρ_CO2 is the density of carbon dioxide at the specified conditions. The other two numbers in the equation are conversion factors to get the correct units of lb/hour. The quantity V₀₂ * RQ is equivalent to the carbon dioxide production rate, V_CO2. The density of carbon dioxide is described as Equation 5 where P is the absolute gas pressure (lb/in²), R_CO2 is the gas constant for CO₂, and T is the absolute temperature (°R). Now that an expression for the density and V_CO2 had been compiled, the metabolic load could be calculated. 

Once the metabolic load and theoretical absorption capacity were defined, it was possible to calculate the theoretical bedlife (TBL). The equation for TBL is described by Nuckols et al. (1996) as Equation 6.