Over the last three years I’ve been working extensively on technical issues related to wood-fired biomass heating systems. The boilers used in these systems burn wood pellets or wood chips. In smaller applications, boilers that burn cordwood using two-stage “gasification” combustion also are used.
Many of these boilers are state-of-the-art products that use microprocessor-controlled air/fuel ratio adjustments to produce average cycle efficiencies of more than 75% and particulate emissions lower than 0.08 lb per million Btu. Both performance metrics are much better than what can be achieved using manually regulated single-stage wood-fired combustions systems.
One of the keys to achieving high efficiency and low emissions is to operate biomass boilers over long on-cycles. In a well-behaved system, the biomass boiler may operate continuously for several hours. It then temporarily shuts down for automatic deashing and to allow some combustion chamber components to cool.
Another operating characteristic of pellet-fired and small chip-fired boilers is their relatively long startup time, (e.g., the time between when the boiler gets a “go” signal from system controls and when it’s ready to go online to deliver heat to the system). From a cold start, a pellet boiler can take 15-20 minutes to go through its startup procedure, which may include automatic tube cleaning, followed by pellet loading, ignition, combustion stabilization and warmup time to bring the thermal mass of the boiler’s steel heat exchanger above the dew point of exhaust gases.
To achieve good performance, it’s vitally important for system designers to create balance-of-system assemblies (e.g., everything other than the boiler) that complement rather than conflict with these operating characteristics. Failure to do this leads to significant reductions in anticipated thermal efficiency and thus higher fuel use. It also leads to higher than expected particulate emissions. Both have been observed in some early-generation systems.
Beyond the boiler
Long on-cycles mandate designers provide a way to handle the significant differences between the rate of heat production at the boiler and the concurrent heat load of the building. This is especially true if the distribution system served by the biomass boiler is heavily zoned, or if the boiler’s heating capacity is equal to or greater than the design load of the building. The latter practice, although common with fossil-fuel boilers, is highly discouraged with biomass boilers.
Sizing biomass boilers in the range of 50-75% of the building’s design heating load helps shift their heat production to “base loading” conditions rather than peak loading. A biomass boiler sized to 60% of building design heating load typically can provide more than 90% of that building’s total seasonal space heating energy needs. This increases to about 96% when the biomass boiler is sized to 75% of building design load. An auxiliary boiler, either new or an existing boiler in a retrofit application, can be set up as a “peaking tool” to supplement the output of the biomass boiler during the coldest weather.
In addition to baseload sizing, water-based thermal storage is used to balance the long on-cycle operating constraint of biomass boilers with the need to carefully control heat flow to a highly zoned distribution system.
Pellet boiler systems often have thermal storage tanks sized between 1 and 2 gal. per 1,000 Btu/hr. of rated boiler output. Thus, a 250,000 Btu/hr. pellet boiler could be coupled with a thermal storage tank in the range of 250 to 500 gal. The lower end of this range is appropriate when the distribution system served by the biomass boiler is minimally zoned and uses higher thermal mass heat emitters such as floor heating. The upper end of this range is appropriate when the distribution system is highly zoned and has low thermal mass heat emitters.
It’s not just volume
Early deployments of pellet boiler systems have demonstrated that just providing a thermal storage volume within the previously noted range by itself is not sufficient to ensure optimal system operation.
The tank also needs to maintain good thermal stratification to optimize the thermodynamic “usefulness” of the heat it contains. Figure 1 provides a simple comparison that illustrates thermodynamic usefulness.
Consider the tank on the left, which has good temperature stratification. The water is at 120º F at the top of the tank and 100º at the bottom. If one assumes an approximately symmetrical temperature gradient from top to bottom, the average water temperature in this tank is about 110º.
Now, consider what happens if vertical flow jets within the tank cause complete mixing, as shown in the right side tank. The water temperature at the top, mid-height and bottom of the tank will be about 110º. Temperature stratification has been destroyed.
Both tanks contain about the same amount of heat, but the tank on the left could still supply a load that might require a temperature of perhaps 118º, 115º or 111º. The fully mixed tank cannot supply any of these requirements. Hence, the higher top region temperature resulting from good stratification increases the usefulness of the energy, albeit only until the tank top temperature drops below the temperature required by the load. Still, it is an advantage that should be employed.
Vertical flow jets caused by vertical piping connections to the tank shell, with no interior flow diffusers, encourage mixing within the tank and thus discourage temperature stratification. This has led some designers to use horizontal piping connections with internal “sparge tube” diffusers, as shown in Figure 2.
A sparge tube is a closed-end internal pipe with an array of holes drilled around some or all its cylindrical surface. The concept is to allow flow to enter the tank at lower flow velocities, and when the entering flow is hotter than the water at the location of the upper sparge tube, to direct that water upward toward the top of the tank. Cooler water entering the lower sparge tube would be directed toward the bottom of the tank.
Although this is an improvement over piping connections that create vertical flow jets, it does not address the situation where the hottest water at the very top of a tank with semi-elliptical end shells needs to be drawn down into the sparge tube and passed on to the load. Due to its reduced density, the hottest water “wants” to stay at the top of the tank. Streamlines in the vicinity of the sparge tube holes do not create sufficient entrainment to draw this hot water down against buoyancy forces. The closer the sparge tubes are to the center of the tank, the lower their ability to access the hottest water at the top or the coolest water at the bottom.
The tank in Figure 3 represents the true proportions of a commercially available ASME pressure vessel sold for hydronic system buffering. It is shown equipped with “straight in” sparge tubes at the sidewall connections. These sparge tubes have holes designed to direct incoming flow upward at the upper tube and downward at the lower tube.
The shortfall with this configuration is the inability of the upper sparge tube to draw the hottest water at the top of the tank downward when it’s needed in the system. Likewise, the lower sparge tube will have little ability to lift the coolest water at the bottom of the tank when that water needs to flow back to the heat source. The end result is less-than-optimal utilization of the tank’s thermal storage potential.
One solution is to create a flow diffuser that allows “access” to the highest and lowest regions of the tank. Figure 4 shows two concepts for this.
Both tanks use a “two-pipe” configuration. This reduces the flow velocity into and out of the tank when there is flow from the boiler and to the load. If the boiler flow rate is higher than the load flow rate, the difference between these flow rates enters the upper tank connection and leaves the bottom connection. If the load flow is greater than the boiler flow, the differences between these flow rates leaves the upper connection and enters the lower connection. Reduced flow velocities entering or leaving the tank reduce mixing and thus help enhance stratification.
Both tanks use vertically oriented sparge tube diffusers located very close to the top and bottom of the semi-elliptical end shells. Sufficient space is allocated for air collection at the top and possible sediment accumulation at the bottom.
All the sparge tubes shown in Figure 4 are bidirectional, allowing flow into or out of the tank. Their position allows the full volume of the tank to participate in flow dynamics, and accesses the hottest and coolest water in the tank. Radial flow from the sparge tubes matches the circulator geometry of the tank shell for even flow distribution.
This concept also can be adapted to two side-by-side tanks as shown in Figure 5.
The flexible connectors shown in figures 4 and 5 are especially important with multiple tank systems. They allow for very slight misalignment of the connections. They also help dampen any vibration carried along by the piping.
I suggest sizing the piping in the vicinity of the tanks for a maximum flow velocity of 2 ft. per second. This creates very little head loss in the piping and allows it, along with the tank shell, to provide good hydraulic separation between multiple variable-speed circulators, as shown in Figure 6.
Thermal storage tanks, especially those that meet the ASME Section VIII pressure vessel code, are expensive. It only makes good sense to maximize their abilities to store heat and provide hydraulic separation.
Here are some more suggestions that will further improve thermal storage tank performance:
- Insulate all thermal storage tanks in biomass boiler systems to R-24 (ºF•hr•ft2/Btu) minimum. Keep in mind these tanks often will be storing very hot water (180º or higher) and that they have large surface areas. Inadequate insulation turns them into uncontrolled “radiators” that overheat the mechanical room.
- Plan any interior diffuser details that allow for air removal at the top and possible sediment accumulation at the bottom.
- Diffuse incoming flows to create gentle horizontal flows inside the tank.
- Don’t destroy temperature stratification with excessive flow velocities due to oversized circulators in either the boiler or distribution portion of the system. When compatible with the boiler design, I suggest using a 30º temperature differential across the boiler and at least a 20º temperature differential across the distribution system under design load conditions.Even higher Delta-Ts are preferred when compatible with the heat emitter design. The higher the temperature difference between supply and return the lower the flow rates, and lower flow rates help preserve temperature stratification. Consider use of variable-speed circulators in both the boiler and distribution portion of the system that reduce flow rates under partial load conditions.
- Use check valves to prevent reverse thermosiphon flows through the tank when no circulators are operating. Why let hot water at the top of the storage tank move backward through the boiler and use the latter as a heat dissipater? Keep in mind motorized flue dampers are not used with biomass boilers due to delayed burnout of residual fuel. Thus, a boiler that’s kept warm by inadvertent thermosiphoning from thermal storage will dissipate heat by convection up the flue, as well as from its jacket.
- Be sure the circulator between the biomass boiler and the thermal storage tank is not running during times when the boiler outlet is cooler than the temperature at the top of the thermal storage tank. This sounds like a “no-brainer,” but I’ve seen it happen more than once. It creates “negative energy flow” between the boiler and tank. Some biomass boilers have internal control logic designed to prevent this. A simple differential temperature controller also can be used to create this logic.
- Specify above-normal insulation on all piping connected to the tanks to reduce standby heat loss.
- Specify tanks with plenty of tappings for temperature sensor wells from top to bottom. These can be used to create boiler control schemes based on temperature stratification within the tank. Remember, it’s easy and inexpensive to plug off tappings (typically 3/4-in. FPT) not used at the time, but usually very difficult to add sensors to measure tank water temperature at locations that weren’t initially planned.
There’s a difference between systems that are “good enough” and those that strive for optimal performance.
It only makes good sense to use the full potential of a thermal storage tank that costs several thousand dollars, requires significant logistics for placement within a mechanical room and provides a vital link between good biomass boiler performance and stable heat delivery.
The details we’ve discussed don’t add a lot to overall system cost, but they definitely improve overall performance.
This article was originally titled “Stranded Btu” in the May 2016 print edition of PM Engineer.