FACILITIES WITH AUTOMATIC STORAGE AND RETRIEVAL SYSTEMS are typically more densely packed compared to warehouses with traditional rack storage. This introduces new fire hazards but also brings certain advantages. What are the fire protection challenges and solutions for robotic top-loading compact storage systems?

The first top-loading compact storage system was introduced to the marketplace about 15 years ago. These systems provide a concept for the rational storage and handling of smaller items or products and handles a given number of plastic storage bins in a three-dimensional structure. The bins are stacked on top of each other in cells of an aluminum profile framework, the top of which carries rails on which battery-powered robots operate. The robots are moving bins and delivering orders to the workstations on the grid-edges where humans pick or refill the items stored.

sow thistle plant
 
The basic design of a toploading compact storage system consists of a grid with cells for the bins stacked on top of each other, battery-powered robots that operate on the top of the grid, and workstations on the grid-edges where humans pick or refill the items stored.

The systems are completely modular, and the number of robots and workstations determine the possible throughput performance. Highly sophisticated logic and traffic algorithms keep track of the items stored and reduce delivery times of bins to the workstations. A small system typically contains 10,000 bins, while the largest systems include 500,000 or more bins.

a chart Large-Scale Control-Mode Sprinkler
 
TABLE 1. Fire Test Conditions and Test Results for Large-Scale Control-Mode Sprinkler Tests Conducted in 2012

Fire Protection Challenges and Advantages

A top-loading compact storage system differs significantly from traditional storage due to its extreme compactness and no aisles. The concentration of combustibles is high, and the open-top bins can also prevent water from overhead sprinklers flowing down through the rack. However, there are certain benefits with the system. The absence of personnel in the storage area can reduce fire ignition hazards associated with humans, the compactness restricts air movement to reduce the fire growth rate, and the grid structure has proven very stable during and after a fire.

In 2009, one of the developers of a top-loading system and RISE Research Institutes of Sweden initiated work to establish fire protection recommendations that meet the needs of end-users and installers. These recommendations are based on a huge number of small, intermediate, and large-scale fire tests, including several large-scale sprinkler test programs. The efforts have resulted in comprehensive fire protection recommendations, offering fire protection solutions for any type of facility and situation.

The fire protection recommendations consist of several different parts and intend to minimize the probability of fire occurrence, fire spread, and exposure to fire. In addition, the recommendations provide guidance on how to design and install a sprinkler system, a smoke and heat ventilation system, a gaseous fire extinguishing system, and an oxygen reduction system. Guidelines include how to facilitate manual fire fighting after a fire has been suppressed or controlled by an automatic sprinkler system. These fire protection recommendations can help risk and warehouse managers optimize the fire protection approach through careful design choices and reduce the costs of fire protection systems.

Sprinkler System Design Challenges

How to provide fire protection with automatic sprinklers is not obvious. Very early, it was recognized that in-rack sprinklers cannot be used as there is no space within the compact storage array to install them. Therefore, any protection needs to be achieved with ceiling-level sprinklers. This raised the concern that the open-top bins may collect water from sprinklers, preventing it from flowing down the storage array and reaching to the seat of a fire. The fact is that NFPA 13 and other sprinkler installation recommendations do not permit sprinklers to protect rack storage with open-top containers.

Two free-burn fire tests were conducted to determine the fire growth rate of a fire initiated at the bottom of one of the narrow flue spaces between two cells, when the grid was completely filled with bins. The bins were loaded with a commodity consisting of plastic cups in cardboard cartons, i.e. the cartoned standard Group A plastic commodity. The tests showed the initial fire growth rate was very slow, and the fire did not accelerate until after approximately 13 minutes. Figure 2 shows the measured heat-release rates.

a chart Fire Test Conditions
 
TABLE 2. Fire Test Conditions and Test Results for the Large-Scale ESFR Sprinkler Tests Conducted in 2017

The fire tests were terminated by the manual operation of a fire sprinkler positioned above the test setup at a fire size of around 8,000 kW and 9,500 kW, respectively.

a chart Total Heat-Release Rate

FIGURE 2. Measured Heat-Release Rates Compared to Standard T2 Fire Growth Rates During Two Free-Burn Fire Tests

Large-Scale Control-Mode Sprinkler Tests

The experience from the free-burn fires tests indicated that protection with ceiling-level sprinklers may be feasible. In 2012, a series of four large-scale, control-mode sprinkler tests were conducted at Underwriters Laboratories, Inc. The tests were conducted with upright, K11.2 (K160), quick-response sprinklers. The following acceptance criteria were established prior to the tests:

  • No more than 10 sprinklers were allowed to operate.
  • The maximum 1-minute average ceiling steel temperature was not allowed to exceed 540°C (1,004°F).
  • No fire spread outside of the center 5 by 6 cells was allowed. Table 1 summarizes the fire test conditions and test results.

The established performance criteria were met in all tests except for Test 2. The fire was controlled as determined visually and by the fact that only one sprinkler activated. However, due to the pipe shadow effect on the sprinkler spray pattern by the sprinkler branch line, the discharge density over the fire area was non-uniform. This allowed for a slow horizontal fire spreading inside the array.

Although the maximum allowed number of sprinklers and the steel temperatures at the ceiling were well below stipulated criteria, the damage criterion was exceeded. Although the open-top bins collected water from the sprinklers, it was shown that sprinklers can control a fire of this type. As the flue spaces are always regular, collapse or leaning of bin stacks across flue spaces is not likely during initial fire development. In addition, the vertical aluminum grid supports limit the possibilities for horizontal fire spread.

ESFR Sprinkler
 
FIGURE 3. Fire Size at Activation of the Sprinkler in the Second Pre-Test Using ESFR Sprinklers, Conducted at RISE in 2016

Large-Scale ESFR Sprinkler Tests

It was questioned whether facilities with higher ceiling heights than the tests described would require another approach. Early Suppression Fast Response (ESFR) ceiling-mounted sprinklers can be used for warehouse protection where in-rack fire sprinklers are requested to be excluded or are simply not possible due to the storage arrangement. These sprinklers deliver large, high-momentum water droplets to penetrate the fire plume. Whereas conventional sprinklers are intended to control a fire, an ESFR sprinkler system is designed to suppress a fire. It should be understood, however, that the fire suppression does not necessarily ensure the fire will be extinguished and means should be provided for manual fire fighting efforts.

In 2016, two ESFR sprinkler pre-tests were conducted at RISE using a single ESFR sprinkler under a suspended ceiling, utilizing a 9-meter (29.5 foot) ceiling height. This resulted in a clearance of 3.7 meters (23 feet) measured from the topmost bin to the ceiling. One test was conducted with a K14.0 (K200) sprinkler and one test with a K22.4 (K320) sprinkler, under as equal test conditions as possible. The fire was suppressed in both tests but was not completely extinguished, since a small residual fire was not discovered until the grid was dismantled in Test 2. Figure 3 shows the fire size at the activation of the sprinkler in Test 2.

As a direct result of the pre-tests, four large-scale ESFR sprinkler verification tests were conducted at Underwriters Laboratories, Inc. in 2017. Tests were conducted at a range of different ceiling heights. As a conservative approach, the sprinklers were installed at the maximum permitted vertical distance from the deflector of the sprinkler to the ceiling. The bins were loaded with the cartoned unexpanded Group A plastic commodity. All cells of the array were filled with bins; however, the cell directly under the robot was only partly filled with bins. This mimics a situation where the robot is unloading bins and provides a more rapid fire growth rate as compared to a completely full grid. The acceptance criteria were identical with those of the control -mode sprinklers tests, except that no more than eight sprinklers were allowed to operate. Table 2 summarizes the fire test conditions and test results.

Performance acceptance criteria were met for all tests. All data in terms of temperatures, fire damage, and numbers of operating sprinklers showed fires were indeed suppressed, although not fully extinguished.

Manual Fire Fighting Tactics and Equipment

Large-scale fire sprinkler tests have shown that fire spread within the array after fire suppression by sprinklers is limited. Typically, fire-damaged bins are found in an area less or equal to 3 by 3 cells.

Any remaining fire after sprinkler system shutoff is small and consists of glowing or flaming combustible material inside deformed bins such that the fire is shielded from water application by the sprinklers. Final manual fire extinguishment of this remaining fire requires fire fighters to approach on the top of the grid. Experience from large-scale fire sprinkler tests demonstrates that damage to the grid structure is very limited and stability of the grid is not a concern, although the area closest to the cells at the starting point of the fire should be avoided. The probability of a full collapse of the grid during or after a fire is extremely low as all cells are interconnected at the bottom as well as the top (i.e. the rails). Despite this, the installation of permanent platforms - either integrated in the grid or installed above the grid - is recommended to facilitate manual fire fighting.

Fire fighting tactics should include the use of smoke and heat ventilation (thermal or mechanical) to improve visibility inside the facility, thermal imaging cameras to identify the seat of a fire, and hose streams of water directed toward the seat of a remaining fire. Improved performance can be achieved with medium-expansion foam, as the foam will fill up any cavities and shielded volumes (Figure 4). Fire tests have shown that fluorine-free foam agents can be used.

It is essential that the sprinkler system is not shut off until direct application of water and/or foam has started. After complete fire extinguishment, the area with damaged cells must be isolated by lifting bins from the cells surrounding this area. This can be accomplished by either a manual bin lift or by robots. The final removal of fire damaged bins requires removal of vertical grid supports in this area.

As an alternative to manual fire fighting operations from fire fighters entering on platforms around or across a grid, a fixed installed system can be used to quench a remaining fire. This system could consist of medium-expansion foam nozzles that apply foam over a certain area or the full area of the top of the grid, depending on its size. It could also be a system with water monitors positioned around the edges of the grid that discharge large volumes of water (or foam) directly on the fire. The monitors could either be remotely controlled or be part of a fully autonomous system that detects, locates, and tracks fires in real-time without human intervention.

Other Types of Fire Protection Systems

For facilities or compartments with a limited volume, gaseous fire extinguishing systems may be used. Even if personnel are not typically present in the actual storage area, carbon dioxide systems (due to the gas toxicity) should be avoided - use inert gas systems instead.

An oxygen reduction system maintains the concentration of oxygen inside a protected space to a level below the specified ignition threshold for combustible materials present. From many aspects, an oxygen reduction system is a suitable fire protection solution. The grid is surrounded by extended walls, personnel are typically not present, and the constant movement of bins (vertically) as well as robots (horizontally) would assist in equalizing oxygen concentration inside the grid. A suspended ceiling above the grid would be desired or, alternatively, walls surrounding the grid extend to the ceiling. The typical maximum design concentration would need to be around 14% volume of oxygen to prevent a fire from occurring, but the desired oxygen concentration should be determined for each installation. Although the bin ports to the grid automatically open and close when bins are delivered to and from the workstations, these openings constitute leakage areas that need to be included in the design of an oxygen reduction system.

foam
FIGURE 4. This illustrates the distribution of medium-expansion foam and its adhesion to the bins during the dismantling of a grid after a sprinkler test. The foam will fill up cavities and shielded volumes and provide extinguishment of any remaining fire.
MAGNUS ARVIDSON

MAGNUS ARVIDSON is with RISE Research Institutes of Sweden. The author would like to thank AutoStore AS for permission to share the results of this work.