The Capitol Hill Riot: A Case Study in Protecting Buildings Against Violent Intrusion
Like many people, I watched the storming of the US Capitol on 6 January unfold live through reporting on TV. And as most, I was horrified to witness the surreal desecration of America’s most sacred symbol of democracy unfolding moment-by-moment. And as a security professional, that horror was amplified even further as I witnessed a cascading series of security failures with full awareness that angry mobs easily turn deadly when group passion supersedes rational judgement.
The next day, as America recovered from its emotional hangover, the leadership of the US Capitol Police was quickly called to reckon. There are obviously many, many questions which need to be answered.
The aim of this article is not to cast judgement about specific matters of security at the Capitol or attribute blame to specific parties. Any statements I could make beyond general critique at this point would be like “playing armchair quarterback without watching the entire game.” No doubt, there will be a comprehensive investigation of the incident and contributing factors which will result in an authoritative report. Rather, our aim in this article is to explore how security measures can be designed to avert similar disaster for the benefit of colleagues protecting other high-risk buildings around the world.
I think we all objectively agree America is a deeply divided nation today. And it is not the purpose of this essay to comment about the social dimensions of this situation. We have many colleagues who sympathize with the grievances of the perpetrators at Capitol Hill. And likewise, we know many others who are equally passionate with an opposing view. Any opinions we hold about politics is irrelevant to this essay and irrelevant to discussion as security professionals. This article will focus strictly on matters of security design. If you, as a reader, wish to share a comment at the end of this article, we respectfully ask that you exercise the same discretion.
Additionally, any statements made in this article regarding failures or vulnerabilities at the US Capitol Building (and examples of hypothetical improvements) are based solely on events as witnessed in the news. Although we have had the honor of serving the Capitol Police as instructors, nothing discussed in this article should be interpreted as a description of security measures present (or not present) at the US Capitol. Our teaching experience with the Capitol Police was over a decade ago and we possess no insider knowledge of the physical security design or current practices at the US Capitol Building.
As many colleagues are aware, a major focus of my work over the past decade has been protecting buildings against risks of armed attack (e.g., terrorist assaults, active shooter situations, etc.). Although there are differences between the dynamics of mob aggression and armed assaults originating in outdoor locations, the measures for protecting buildings and occupants against both threat scenarios are conceptually similar. Both situations require having an effective lockdown procedure supported by physical security design and building infrastructure.
Securing buildings against incidents such as the Capitol Hill Riot begins at the first protective layer
Regardless of the location and complexity of a facility, the first protective layer will always be regarded as the exterior perimeter of the facility grounds. In an ideal situation, the outdoor perimeter would be designed as a contiguous protective layer using intrusion detection systems and anti-personnel barriers (e.g., fencing, walls, GPBTO, etc.) capable of providing enough delay for a response force to intervene before adversaries can penetrate into occupied building structures. However, withstanding locations like military bases, airports, prisons, and critical infrastructure sites, rarely do we have the benefit of employing robust perimeter protection due to various restricting factors (e.g., site location, requirements for public accessibility, cost-benefit justification, etc.).
Nevertheless, even in the absence of perimeter barriers, early and reliable detection of an approaching attack at the first protective layer is critical in implementing an effective lockdown plan. At the US Capitol Building, this would have been the role of visual observation and reporting by officers at the outdoor barricades and security operations center personnel monitoring events by CCTV. The moment those barricades began to fail or the crowd’s intentions to penetrate the barricade became obvious, a lockdown should have been initiated.
In cases where we don’t have the benefit of officers posted outdoors, other measures can often be used to improve early detection and speed the lockdown process. For instance, employing gunshot detection sensors can provide immediate warning when an armed attack commences outdoors. Armed attacks by outsider adversaries often initiate against people located outdoors before attackers enter buildings. Some examples include assaults at the First Baptist Church (Southerland Springs, 2017), Burnette Chapel Church of Christ (Antioch, 2017), Inland Regional Center (San Bernardino, 2015), Curtis Culwell Center (Gardland, 2015), Charlie Hebdo office (Paris, 2015), Bardo National Museum (Tunis, 2015), and the Ozar Hatorah School (Toulouse, 2012). In these types of situations, outdoor gunshot detection sensors can provide critical seconds necessary for initiating a lockdown response.
In other cases, positioning panic buttons capable of automating the lockdown process and alert notification in locations where employees are positioned at reception desks or outdoors (e.g., guardhouses, parking attendant sheds, etc.) can also expedite the lockdown process. This is a common measure I recommend when working with large school campuses where shared buildings (e.g., cafeterias, media centers, etc.) are normally kept in an unlocked state. Other situations where I’ve made this recommendation include resort hotels in response to situations such as the 2015 Sousse attacks.
To ensure that a building lockdown protocol can be implemented effectively, all exterior doors and glazing should be intrusion-resistant and capable of providing reasonable delay. Outward-swinging commercial steel doors and solid-core wooden doors with appropriate locking hardware will usually provide 120 seconds or more of delay against forced entry without the use of optimal tools. That often provides sufficient time to facilitate the arrival of an on-site response force. As a consultant, rarely do I advise clients to replace or upgrade exterior steel and solid-core wooden doors.
Glazing, however, is a different matter. One of the most common problems I encounter when assessing buildings against violent forced entry threats is window glazing and doors constructed of annealed or tempered glass. Tempered safety glass provides very minimal delay against forced intrusion. Annealed glass, even less. According to testing documented by Sandia, 0.25 inch tempered glass provides 3-9 seconds of delay against an intruder using a fire axe and the mean delay time for penetrating 1/8″ tempered glass with a hammer is 30 seconds. Furthermore, testing documented by Sandia does not account for the weakness of tempered glass after first being penetrated by firearm projectile. In our penetration tests of 1/4-inch tempered glass windows using shots from a 9mm handgun prior to entry, delay time was only 10 seconds.
If we already have existing annealed or tempered glass windows, upgrading windows with the use of anti-shatter film properly secured to the frame can upgrade the delay time performance to 60-90 seconds depending on the adversary’s tools. In many facilities, anti-shatter film is an inexpensive upgrade and sufficient when used in combination with other measures. However, in situations where the risk level is high and/or adversaries may be equipped with optimal tools, glazing rated under standards such as ASTM F1233-08, EN 1627, or even UL 972 would be justified.
Although I am unaware of the window specifications at the US Capitol Building, photos and videos suggest outdoor glazing included a mix of annealed and tempered glass windows reinforced with the use of anti-shatter film. And there are at least three clips documenting protestors using objects such as riot control shields and batons for penetration. I’m unable to cite the delay performance of those windows since the clips I’ve seen didn’t display when penetration first commenced. What we can observe, however, are cases where the glass was retained by the film membrane as a single unit and tearing at the edges of the frame. This type of failure often results from improper frame attachment. To avoid this problem, specifications for anti-shatter film should always require mechanical or cement bond frame attachment.
Even better performance could have been achieved with the use of laminated glass. Laminated glass is a composite material constructed of two or more layers of glass bonded to a PVB or polycarbonate interlayer. According to Sandia’s test data, 1/4-inch laminated glass provides 18-54 seconds of delay against forced entry by fire axe and the mean delay time for penetrating 9/16-inch laminated glass is approximately 1.5 minutes by hand tools. Most glazing products tested and rated under forced entry standards are constructed of laminated glass.
At the US Capitol Building, it is likely anti-shatter film was installed on exterior windows as a blast mitigation measure. When laminated glass fails under overpressure loading from explosive blast, it typically remains as a solid, high-mass object when it’s projected into the building, quite dangerous to anyone nearby. Anti-shatter film with a proper catch mechanism fails much more safely during IED attacks. However, after comparing all ‘pro’s and con’s,’ laminated glass is usually my preferred choice when addressing first-floor windows in unoccupied areas due to its greater forced entry resistance.
Secure Entrance Design
Although intrusion-resistance of the exterior façade is important, it is only valuable if entrances are secured when an outdoor attack is detected. In high-traffic public buildings, a first measure should be securing normally unlocked public entrance areas (e.g., lobbies, etc.) as independent protective layers using full-height, intrusion-resistant walls and turnstiles constructed of laminated glass.
Many government buildings I’ve assessed omit this measure due to the presence of police or security officers in lobbies. However, the presence of officers in public lobbies is no substitute for barriers. If officers are caught by surprise or confronted by a large group of aggressors, physical barriers are a crucial line of defense. The 2014 attack at Parliament Hill in Ottowa clearly demonstrates this concern. After wounding an unarmed officer at the Centre Block building entrance doors, gunman Michael Zehaf-Bibeau was shot by another officer located nearby and able to run successfully through the Rotunda and Hall of Honour corridor past unlocked rooms occupied by members of Parliament before being contained and neutralized. Had the entrance design at the Centre Block building employed a secure layer of full-height intrusion resistant barriers (e.g., walls, turnstiles, etc.), it is not likely Zehaf-Bibeau would have made it past the lobby.
The following before-and-after images illustrate how we approached this matter for a commercial client a few years ago with similar threat concerns. The height is less than ideal for situations such as a mob invasion, but demonstrates the concept.
In buildings where it is impractical to design a secured public reception point, it is especially critical that normally-unlocked exterior doors can be secured quickly and reliably. The only way to ensure this capability is through the use of electrified locking systems and a lockdown macro programmed using access control software. Reliance on people to manually lock exterior doors is a recipe for disaster in these types of situations. Ideally, the access control system should be designed so that one button activated by a control room operator (or perhaps a receptionist) automatically locks all exterior doors, lobby turnstiles, and interior doors strategically located to further delay movement inside the building.
And when selecting electrified locks, avoid the use of electromagnetic locks due to concerns about fail-safe operation during fire alarms, push-to-exit button and sensor requirements, and other problems. For reasons of expediency, I will skip further discussion about this subject. I have written quite a bit about this matter elsewhere and encourage you to check out other articles for more information regarding electrified locks and egress issues during attack events.
Deployable Indoor Delay
In high-risk facilities, added delay can be achieved with the use of additional hallway doors and turnstiles positioned between likely entry points and areas where high-valued assets are located. In many cases, these additional doors can be maintained in an open state using fire door release hardware to avoid inconvenience and deployed (released) when a lockdown event is activated through the access control system.
Although this measure can be quite effective in creating added delay time, choosing the location for deployable security doors often requires careful assessment of likely adversary movement routes and egress paths to avoid obstructing escape by building occupants.
Dispensable Indoor Delay
In addition to structural barriers, there are several types of dispensable barrier systems that can be installed in buildings designed to disorient or impair intruders when activated. Some examples include rigid polyurethane foam, sticky foams, stabilized aqueous foam, and systems for dispensing fog and aerosolized chemical irritants.
To date, the use of dispensable barrier systems has largely been limited to high risk environments with low occupancy, such as nuclear facilities. In a building such as the US Capitol, the only option I might consider practical are systems designed to dispense chemical irritants. By comparison to police officers deploying chemical sprays or munitions, installed dispenser systems are much more efficient for disseminating riot control agents in large indoor areas. If such a measure is considered, I recommend limiting dispenser nozzles to locations along most likely adversary movement paths (such as main corridors near public entrance doors) and away from areas normally occupied by employees, VIPs, and critical egress paths.
High-Risk Assembly Areas
In high risk buildings, such as parliamentary buildings and the US Capitol, areas where VIPs assemble en masse should be secured by a robust protective layer even if there are provisions for moving VIPs to a nearby safe haven. As demonstrated in the 2014 Parliament Hill attack, armed attacks happen very quickly and there may be insufficient time to move VIPs to a safe location.
During the January 6 attack, I was mortified to witness how easily the rioters penetrated into the Senate and House chambers. The vulnerability of Congressional chamber doors prompted staff to barricade the doors with furniture with armed officers in over watch. Furthermore, news reports suggest it took as much as one minute for Capitol Police officers to manually lock all doors to the Senate chamber. In fairness to the US Capitol, a very similar situation occurred in Ottowa in 2014.
To avoid this critical vulnerability, doors to high risk assembly areas should be equipped with electrified locks operated automatically by the access control system when a lockdown macro is activated. Additionally, doors to high risk assembly areas should ideally be specified for forced entry resistance against group attack.
The best standards for this purpose are the US State Department’s SD-STD-01.01 and ASTM F3038-14. Both standards were designed exactly with this type of situation in mind and are the gold standard for specifying safe room doors.
The SD-STD-01.01 test protocol (for complete DOS certification) involves a series of ballistic tests against different parts of the specimen (shotgun, 5.56mm, and 7.62 NATO) and forced entry tests involving a team of aggressors conducting a series of attacks against the specimen at different parts with a diverse toolset. The tools and number of active test personnel varies based on time of test. Specimens are rated according to their timed forced entry-resistance against three attack levels: Five minutes (two test personnel), Fifteen minutes (six test personnel and larger range of tools), or Sixty minutes (six test personnel and greatest range of tools).
The following video demonstrates the SD-STD-01.01 test protocol.
ASTM F3038-14 is similar to SD-STD-01.01, but with some differences regarding number of attackers, ballistic resistance testing, and rating scale levels. ASTM’s testing approach involves six persons conducting a series of aggressive attacks against the barrier specimen with the use of various tools. Different parts of the barrier are subjected to independent timed tests. Specimens are rated according to their timed resistance against four levels of attack: Five minutes, Fifteen minutes, Thirty minutes, or Sixty minutes.
Although these types of doors would have ensured much better performance than witnessed at Capitol Hill, retrofitting multi-door entrances with doors rated under SD-STD-01.01 and ASTM F3038-14 may require some architectural creativity due to single leaf construction. Nevertheless, when protecting assemblies of Congress and parliament, the level of risk more than justifies the effort and expense.
Safe Rooms & Safe Havens
It is reported that all members of Congress were moved to a sublevel safe haven before the mob penetrated into Congressional chambers. At least this part of the security plan, “The Alamo” as it were, seemed successful.
However, I have encountered similar buildings in other nations that did not have a safe haven sufficient for large groups of VIPs. If that is the case, addressing this vulnerability should be a top priority. Rather than describe measures for designing a safe haven for group occupancy, I refer readers to US DoD’s UFC 4-023-10 “Safe Havens” as the best publicly available resource on the subject.
In addition to the issue of protecting groups of VIPs moved from assembly chambers, VIP offices should be designed to function as temporary safe rooms in the event an attack occurs when Congress or parliament is not assembled and for the benefit of staff members.
Although I have no idea of how VIP offices in the US Capitol are constructed and secured, the fact that rioters held a frat party in Nancy Pelosi’s office suggests there was a major failure of some type.
In cases where it is impractical to upgrade all VIP offices to a high-level of forced entry resistance, I recommend that offices at least be upgraded to provide a minimum of 45 seconds of delay against aggressors using impact force and firearms as entry tools. Although 45 seconds isn’t much time against a committed adversary, many attackers will abandon the effort in favor of easier targets. Five minutes or greater is more preferred if budget permits.
Following are some examples of how this issue can be approached.
Public Address Infrastructure
Lockdown procedures should always include an audible alert announcement to ensure that all building occupants are aware when an attack is in progress and expedite personal response. Although it appears most members of Congress and police were quickly aware when the building lockdown was ordered, several news reports suggest the first warning many occupants received was when they heard nearby commotion and gunfire. Additionally, in none of video footage I’ve watched can an audible announcement be heard from the building’s public address system.
This is not a light matter. In the Capitol Hill riot, we were fortunate that members of Congress were assembled near Capitol Police officers in radio communications with the security operations center. If such an event occurred under other circumstances, many could have been delayed in initiating protective actions.
If we are responsible for protecting a building at risk of attack (armed assault or mob violence), effective public address capability should be a top priority. To facilitate rapid alerts, security control rooms receiving event reports or assessing alarms & CCTV should have direct access to the building’s public address system. There should also be adequate speakers available throughout the facility to ensure good audibility.
In addition to public address announcements, high risk organizations should always employ an electronic mass notification system (MNS) as a redundant communication system and to warn those who may not have heard audible announcements. And when important developments occur, updates can be issued to employees as follow up messages. Circumstances warranting updates may include notification when police are clearing the building or if new dangerous developments occur, such as a building fire while people are sheltered in refuge.
There are numerous mass notification systems on the commercial market suitable for this basic purpose. Most systems permit messages to be issued by SMS text, email, and push data transfer through Android and iOS applications. For organizations with deficient public address systems or who desire an integrated panic alert system, our company, CIS, spearheaded the design of GuardianCall which automates all key notifications with one button press using a body-worn pendant device.
Response Force Intervention
Last and not least, an armed response force must intervene quickly and effectively when an attack is detected. Even the best designed Physical Protection System (PPS) will fail if the responding security force is delayed or ineffective in their actions.
As for the response at Capitol Hill, I don’t know where to begin in comment. It’s a subject far beyond the scope of this article and I’ll reserve critique until more details become clear.
From a security perspective, the events of January 6 were clearly preventable. Yet mitigating such situations does require a carefully integrated approach to physical security design and response preparations. Although public attention is currently spotlighting problems at the US Capitol, I’ve witnessed identical vulnerabilities in many facilities of similar stature around the globe. The potential conditions contributing to security failure on January 6th are not isolated cases.
As security professionals, there are always unique circumstances which influence the practicality and risk justification of security measures in different environments. Yet the principles outlined in this article are universal and can be scaled and adapted with artful application to any facility including government and corporate offices, historical sites, schools & universities, community centers, hospitality facilities, event centers, churches, and more.