Robert Fulmer Harraseeket Restoration Portsmouth, New Hampshire ABSTRACT This paper provides an overview and knowledge base to properly evaluate slate roof systems for maintenance requirements, system longevity determination, and water infiltration issues. Beginning with a history of the slate roofing industry in Europe and U.S., discussion then evolves into the composition, characteristics, and identification of the various categories and types of slate. An examination of the various applications and characteristics of slate roofing are followed by an in-depth discussion of proper installation techniques in accordance with industry standards and “best trade prac¬ tices.” Critical individual elements are then discussed, such as flashing details and materials, the use of modem materials and technologies in a contemporary slate roof system, the specification and comparison of domestic and foreign slate, simulated slate products, water infiltration issues, etc. The proper procedures for conducting an “intru¬ sive” slate system inspection and how to analyze and interpret the data into a useful format for making recommendations on the maintenance, restoration or replacement are provided. It concludes with a demonstration of how to format data into useful long term maintenance projections and plans for the client. Robert L. Fulmer Robert L. Fulmer is a licensed Historic Roofing Contractor with the state of Massachusetts, a historic preservation consultant, and president of Harraseeket Restoration, Inc. in Portsmouth, New Hampshire. He is the fourth generation in the roofing business, performing restoration services to the exteriors of historic properties. This includes installation and maintenance services to architects, historic preservation organi¬ zations, churches, institutions, and private owners on an international level. Harraseeket Restoration has been the recipient of numerous awards for excellence in preservation and featured in national media. As the author of articles on exterior preservation issues, Fulmer also provides lecture services and expert wit¬ ness testimony. He is also a member of RCL Fulmer— 13
Slate roofs last forever, right? How many times as roofing professionals have we heard this myth perpetuat¬ ed? Occasionally, even members of our own profession do it. While longevity is certainly one of the significant advantages of a slate roof system, in reality, “nothing lasts forever” – slate roofs included. Slate roofs can and do fail – usually for one of two reasons – either by natu¬ rally reaching the end of their service life, or by an event that precipitates a premature failure. But what is the “service life” of a slate roof? What could possibly cause a system designed to last in excess of one hundred years to fail prematurely? How do we determine the causes and report on them effectively? As consultants, it is imperative that we be able to answer these questions and more. To get there, however, we need some basic facts and information about these unique roof systems. Slate roofing and its adjunct, coppersmithing, are centuries-old crafts whose techniques are, in North America for the most part, passed from one generation to the next. Although a recognized and struc¬ tured apprenticeship approach exists in some European countries, slate roofing remains a “niche trade” in the U.S. Whether considered “obscure” or “unusual,” the fact is, as a complex trade, there are a number of consid¬ erations, characteristics, and installation procedures that are specific to slate systems. Over the past one hundred years, until relatively recently, there has been surprising¬ ly little written about these systems in North America. While it is not necessary to become “master slaters,” our goal as consultants is to obtain enough information to properly evaluate and diagnose these unique systems, then format our data into accurate and useful reports. ROOFING SLATE DEFINED A roofing slate is a rock – hence the “lasts forever” myth. More specifically, slate is cut from sedimentary rock extracted from the upper strata of the earth’s crust. Formed under intense pressure from deposits of sediment thousands of years old, the sedimentary rock contains the minerals specific to the area of origin. These minerals are what give the slate its coloration and quality. Each deposit or “bed” may contain other geological debris, such as mica or quartz, that combine to create a local color or slate type. Approximately twelve primary slate colors plus local variations are quarried in Vermont and eastern New York. Since the late 1700s, this region has consistently produced the majority of roofing slate quar¬ ried in the United States. Other regions that have made significant contributions to the industry are Maine (Monson quarries), Connecticut (small, family-owned quarries), Pennsylvania (Peach Bottom slate), and Virginia (Buckingham slate), to name the more well known. The principles of the quarrying process have changed relatively little over centuries. Yes, sophisticated explo¬ sives have replaced black powder for removing the large “blocks” from the quarry pit wall, and huge dump trucks have replaced mules and steam-driven “conveyors” to get the block to the mill, but it is still dangerous and laborintensive work. Once at the mill, the rough blocks are evaluated for purity, color, uniformity, and grain, then “rough sized” with a diamond blade saw. Then the blocks are split into slate of the prescribed thickness trimmed to their finish size, “mill punched” with two nail holes, and palletized. In many quarries, the splitting, punching, and palletizing are performed manually or on equipment that was used at the turn of the century. Only fifteen percent of the final product of this laborious process finds its way onto a roof. Eighty five percent of the slate that comes out of the quarry pit is considered waste. For the consultant, the most critical information obtained from knowing the slate composition and quarry¬ ing process is to be able to determine slate type and cate¬ gory. Each slate type has its own geological make up that is unique to its area of origin, and in some cases, the quarry from which it was produced. The word “type” in this case refers to individual slate colors such as “Vermont Gray,” “Unfading Red,” “Weathering Sea Green,” etc. Another key piece of information contained in the name of the slate is its weathering characteristic. All slate changes slightly in color when exposed to the weather. Some slate types change color completely. Those slate whose color changes slightly are classified as “unfad¬ ing.” Slate whose color is changed substantially is classi¬ fied as “weathering.” Those in between are “semi-weath¬ ering.” The slate type and weathering category give a general indication of porosity and hardness, which in turn define service life for a slate system. Simply put, Fulmer— 15 slate that is softer absorbs more water. It then begins a deterioration process that results in a shorter service life than a denser or harder type of slate. This hardness is a result of the depth (or age) of the slate bed and its miner¬ al composition. It is true, then, that all slate are not creat¬ ed equal. To further assist in determining the service life of a particular type of slate, refer to the American Society for Testing and Minerals (ASTM) test results for that type. In 1957, ASTM Standard C406 was developed as a stan¬ dard for grading roofing slate. Three tests were then developed. Test C120 is “Modulus of Rupture” in P.S.I. process. Test C121 determines water absorption capacity, and Test C217 is “Depth of Softening” – a function of hardness. Based on these test results, ASTM C406 groups the slate types into three groups. Grade SI has a service life greater than 75 years. Grade S2 has a service life of 40 to 75 years. Grade S3 is the least desirable, with a service life of 20 to 40 years. Being able to identify the slate type and category to determine service life becomes a critical first step when recommending repairs vs. replacement or maintenance projections to a client. THE INSPECTION PROCESS When conducting a thorough slate roof inspection, it’s helpful to adopt the mindset of a good police detec¬ tive. The investigative processes are very similar. First, establish the “corpus delecti” or the purpose and scope of the consultation. Some of the more common requests are from clients who want to establish system condition for an effective maintenance or replacement program based on remaining service life. Or, in some cases, the roof sys¬ tem has failed prematurely and the cause must be estab¬ lished. Far too often, requests come from clients planning litigation whose slate systems have failed due to non¬ standard installation techniques. Whatever the reason for the consultation, the approach is basically the same. PREPARING THE CASE FOR “COURT” A good detective begins his work by studying all available facts objectively. As he proceeds, he utilizes all available tools at his disposal. For the consultant, the most valuable “tool” to establish system condition and verify the materials and installation techniques utilized, is the “intrusive inspection.” This technique involves get¬ ting up on the roof and removing sections of the slate. These “test areas,” as they are known, are not simply removing a slate or two here and there, but are part of a documented, systematic approach that makes the intru¬ sive investigation so effective. In essence, a consultant is “doing the forensics” – documenting the system in such a way that a fair repre¬ sentation of the materials used, their condition, and the installation techniques utilized by the workmen are deter¬ mined. Be methodical! It is essential to have a plan in place before the inspection begins. Assemble and review all available construction documents and building histories. Old plans, specifications, contractor’s records, etc., can sometimes be obtained from the client. Building histories and photographs may be available from the local histori¬ cal society or town offices. Knowing the age of an older building is very helpful. In the case of newer construc¬ tion, the architect or engineers are excellent sources. Use this information to plan a approach. For example, is this slate roof the original or a replacement? Old photos may verify this. Remember, determining the age of the slate is essential in determining the service life. Are there current leaks in the system? By carefully pinpointing the source of water infiltration, test areas may be located in their vicinity. This is a valuable tool in documenting system failure and determining whether natural deterioration or poor workmanship is the culprit. How will the roof be accessed? Can it be walked? Is fall protection and a “chicken ladder” sufficient, or are scaf¬ folding or a lift required? As these questions are answered, the plan takes shape. It’s time then to consider what is arguably the most important part of the inspection – the documentation. Digital cameras and digital video are perfect for the roof consultant. 35mm cameras work well also. Whatever for¬ mat is selected, remember: You may or may not know it now, but you could be literally “preparing your case for court.” Few clients and fewer juries will venture out onto a slate roof, so take lots of photos and supplement them with video if possible. The equipment used should have a high enough reso¬ lution to produce good quality results when enlarged. One reason I prefer my digital camera is that the photos are easily viewable on a laptop computer for the client, but more importantly, they can be formatted into a Powerpoint® program. When testifying as an expert wit¬ ness in court, the jury may not know a slate hook from a ripper. However, there are few things more compelling than a large format Powerpoint* presentation to educate the jury (or client) and make a case. Be prolific with the photos. Take before, during, and after photos of the inspection and the test areas. Label the points to be made by writing a caption on masking Fulmer— 16 tape or white paper and attaching it before the photo. This is useful for pointing out deteriorated flashing details, broken solder joints, number and location of the test areas, etc. After taking the photos, make a back-up video of the same thing with a running dialogue of what is being filmed. Education of people not familiar with slate roofing is a big part of what the consultant must do; these tools make the job easier and increase our cred¬ itability. Before beginning each inspection, check the “inspec¬ tion kit.” This handy item is a canvas bag that contains all essential items to do the job. In it, pack a body har¬ ness and lifeline, a “loaded” nail pouch, a slate hammer, ripper, and stake. Also carry snips, a small pry bar, a flashlight, and caulking gun. For materials, take copper nails, slate hooks, small pieces of copper flashing, and bituthene. It sound like a lot, but it is manageable (and essential) on the roof. Now that the plan is in place and the tools are assem¬ bled, it’s time to go to work. As discussed earlier, the first step of the inspection is to determine the service life of the slate. To do this, the type, age, and condition of the slate must be identified. You may be able to recognize most slate types. If not, contact several quarries or slate brokers from each region and request their sample kits. In the case of weathering slate, it is helpful to build an inventory of samples of various known ages. These samples can then be com¬ pared to the slate on the subject property. Once the slate type has been identified, the average life or “service life” of that slate system may be determined. But how old is the slate? If the building’s age is known and the existing roof is original, then it’s simple. If unsure, it may be necessary to remove a test area of slate to inspect the sheathing for evidence of another set of nail holes. Don’t rely on view¬ ing the backside of the sheathing in the attic, as the slaters may not have used nails long enough to penetrate the sheathing. Accurately dating the slate is not always an exact science. In any case, the most important obser¬ vation is of the slate’s condition. As slate begins to dete¬ riorate, three things usually occur in sequence. First, “shaling” or flaking begins to occur on the backside of the slate. This “powder” can be easily brushed off and indicates an increase in water absorption. Following this, the slate can begin to “delaminate” or separate into thin layers, usually around the edges. In the final stages, the slate becomes soft and can be crumbled in the hand. The slater’s practice of “ringing” the slate (holding the slate between two fingers on one comer and tapping lightly with a slate hammer on the opposite corner) is an excel¬ lent means of determining the level of softness. The slate should “ring” when tapped, not emit a soft “thud.” After establishing the service life of the roof, proceed by inspecting each component of that system beginning with the structural integrity of the support system. Evaluate rafter span, size, spacing, sheathing type, and thickness to ensure the substrate is adequate to support the weight of a slate roof (generally seven to eight pounds per square foot). If, after consulting the structural span and download tables, there is uncertainty about the integrity of the system, an engineer should be consulted. Next, begin removing slate for test areas. The number and size of these areas varies with the needs of each pro¬ ject, but should be sufficient to provide a representative indication of conditions as well as problem areas. Once down to the substrate, the underlayment should be identified. Thirty-pound felt or “slater’s felt” are most common, but bituthene or other materials may be encountered as well. Due to heat transfer from the slate and attic, most thirty-pound felt begins to break down in approximately 40 to 50 years – usually turning to “dust” by 60 or 70 years. Next, the fasteners should be identified. Copper “slater’s nails” are the preferred nail, but steel or galva¬ nized nails appear as well. A lot of generalization has been circulated about the historic use of steel nails. Even in the professional community, the belief that any steel nails in a slate roof will fail prematurely still exists. The reality is that in the 19th and 20th centuries, there were a number of grades of steel used in nails. The better grades will last the life of the slate. The important thing is the size of the nail head. Too small, and the nails will enlarge the hole in the slate with wind-induced vibration, causing slippage. The condition of the nail shank at the point where the slate makes contact is also important when steel nails are used. Moisture present in the slate can oxi¬ dize softer steel at the point of contact, causing the shank to fail over a period of years. This corrosion point is visi¬ ble just below the head of the nail made of softer steel. The last critical component of the inspection process is the flashing details. The fact is, flashing details don’t last as long as the slate. In the case of the harder slate types, one system can go through as many as three or four sets of flashing during the service life of the slate. Flashing failure – whether premature or through natural deterioration – is the leading cause of water infiltration in a slate system. Unfortunately, over the years, many slate Fulmer— 17 roofs have been replaced when this fact has been misin¬ terpreted by owners or industry professionals unfamiliar with slate systems and flashing replacement techniques. If flashing problems are suspected, two determinations must be made. First, the flashing material must be identi¬ fied; then, it must be determined whether natural deterio¬ ration or premature failure from poor installation tech¬ niques is the cause of the failure. Copper is the most common flashing material in a slate system. Its malleability and long service life make copper a good candidate. Unlike other sheet metal, cop¬ per thickness is designated by the “ounce” rather than “mil” measurement. Sixteen-ounce copper simply means that one square foot of the material weighs 16 ounces. Copper is available in a number of different weights, with 16- and 20-ounce being the most common. Certainly, other metals, such as lead, zinc, and galva¬ nized steel, are used as flashing material. But copper in its natural or lead coated form is the traditional choice. When inspecting flashing details, begin by determin¬ ing approximate age and condition. Age is less relevant than the location of the flashing (its exposure) within the roof system. For example, a 16-ounce copper ridge cap could last 70 or 80 years due to the fact that very little water is coursing over it. The same 16-ounce copper may only last 40 to 50 years as a valley in the same system. Look for the obvious at first, such as holes worn through the copper in high volume drainage locations such as val¬ leys, ice belts, and gutters. The “drip line” is that area of a flashing along the edge of the slate or other materials that allows water to drip onto the flashing details. Most commonly, these areas are below the edge of slate over¬ hanging a valley, gutter, etc. It is along this drip line most flashing wear through first. In copper, tiny pinholes appear, then enlarge over time. A good indication that copper is about to wear through is the presence of “streaking.” The green patina on copper is an oxide. The oxidization process is contin¬ uous and protects the copper throughout its life. When the material becomes very thin, it can no longer support the oxidization (green patina) and streaks of natural cop¬ per begin to show through the patina. Another indication of problem areas is the use of asphalt cement or caulking as a temporary fix on flashing. In the case of copper, industry professionals cannot agree on the extent to which asphalt degrades copper. The reactivity that occurs between these two materials is attributed to the sulfur present in the asphalt. As water courses over the asphalt patch and mixes with the sulfur, a diluted form of sulfu¬ ric acid is created. This acid “cleans” the patina off the copper, and streaks of natural copper appear. It is diffi¬ cult to determine the extent to which this shortens the life of the flashing. We do know, however, that over a period of years, the copper becomes thinner where it contacts the asphalt in an open location exposed to moisture. The second principal reason for flashing failure is poor installation or “non-standard trade practices” result¬ ing in premature system failure – “premature” in the sense that water infiltration occurs before the flashing has deteriorated naturally. Of all the possible installation errors, two are most common by far. The first is poor soldering technique, resulting in failed solder joints. The second is failing to allow for thermal elongation (or contraction and expan¬ sion) of the metal. All sheet metal exposed to the ele¬ ments contracts and expands with daily fluctuation in temperature. If this movement is inhibited, failure will occur at the weakest point. This could be a poorly sol¬ dered joint or an area of multiple bends in the metal. “Face nailing,” or exposed fasteners through the metal, are an indication of a problem as thermal movement of the materials cannot occur. Other common examples of improper installation techniques are the absence or improper location of expansion joints. The use of oversized panels in a flat seam or standing seam configuration creates excessive “oil canning” and prematurely fatigues the material. This is evidenced by failed solder joints or hairline splits in the material at bends or other high-stress areas. Revere Copper’s Copper and Commonsense and SMACNA’s Sheet Metal Manual are excellent references for tables that determine expansion joint requirements, specify proper panel sizes and configurations, etc. Again, document observations well, because the determination that a system has failed prematurely is one of the greatest responsibilities a roof consultant can encounter. After the flashing system, the typical inspection is now complete. However, no discussion of roof inspection is complete without mentioning safety. Most slate roofs are steep and many are difficult to access. Personal safety should always be one’s greatest priority. Integrate the safety issues into an inspection plan. Explain to the client why a lift or staging set is required. Always carry a body harness and lifeline in the inspection kit and wear it. Resist the urge to “just run up there” to take one more measurement or photo. On most slate roofs, a mistake could lead to serious consequences. Fulmer— 18 REPORT WRITING TECHNIQUES Once all information has been gathered and docu¬ mentation made, it’s time to format the data into a useful and professional document. The written report, like the inspection, should be methodical and thorough. The con¬ tent must be meticulously accurate and presented in a format that is both easy to understand and access. The report should consist of separate sections that flow logi¬ cally in sequence. Each report may be slightly different, depending on the client’s needs and purpose. What fol¬ lows is an explanation of a typical report format contain¬ ing all of the basic sections listed in their logical order. The Report Cover: The report should be bound in a hard or soft cover. The front cover should contain the title of the report (i.e., “Water Infiltration Analysis for #161 State Street”), who it is prepared for and by, with the date (i.e., “Prepared for the State Street Owners’ Association by Robert Fulmer of XYZ Associates on September 1,2003”). If possible, a photo of the subject property should appear on the cover. Immediately inside the report cover (or on the binding), the company name and contact information should appear. This configura¬ tion gives the client a clear idea of what the document is and the consultant’s contact information, without search¬ ing- Table of Contents: The first page should be the table of contents. It should be the “map” for the report, listing sections by page number. It is also helpful to tab each section with dividers for quick reference. Building History: If there is sufficient documenta¬ tion, a brief building history is both interesting and help¬ ful. Include any major construction or renovation events to provide a clear idea of system dates and chronologies. Inspection Techniques: This section should describe the exact process used to arrive at the informa¬ tion obtained. List the number and location of test areas. Also acknowledge any paper documents or individuals from whom information was obtained. Describe the means of access to the roof. Lastly, be sure to include information regarding photo formats (video, 35mm, etc.) and whether it was possible to obtain actual samples of the materials. This section is particularly helpful to a client’s attorney if the case is going to trial. Roof Plan: Include a roof plan with areas marked to correlate with information in the report. There are a num¬ ber of CAD type software programs to make this simpler. Existing Conditions: This section should be a writ¬ ten “snapshot” of the slate system. It is a thorough cata¬ loguing of all the components of the system and their existing condition. Structural details and condition should be discussed here, along with slate type and age, etc. Any deterioration should also be noted here. Be objective; this section is simply a statement of facts about all of the materials and their conditions that com¬ prise the system. Be as complete as possible about mate¬ rials identification and age. Recommendations and Conclusion: You’ve seen the patient, now write the diagnosis. This is the section to describe the consultant’s interpretation of the data. Reiterate the inspection process, describe how the infor¬ mation was obtained and interpreted. Support conclu¬ sions with as many sources as possible. Use tables, graphs, trade publications, etc. If there is an installation issue, use professional documents to define “standard trade practices.” Some useful sources are NRCA and SMACNA manuals, Joe Jenkins’ book, The Slate Roof Bible (Editor’s Note: This is available through RCI’s Publications List), and the 1926 National Slate Association publication, Slate Roofing, to name a few. This is the section to show one’s expertise. This is why the client hired you. Be aware though, if in doubt about something, say so. Only state the facts known to be accurate. Don’t try to bluff your way around an issue about which you are unsure. Remember, nobody knows it all and if the case goes to trial, the other side’s attorney will dissect your report with surgical precision to find errors or inconsistencies. Equally as important, remember that in order to be professional, a consultant must always remain objective. A consultation report is not a sales tool or solicitation. If your company also performs the type of work you are consulting on, consider including a letter of disqualification from bidding on future work related to the project in the report. Project Cost Estimate: Most clients will want cost figures for budget purposes. Discuss the details with the client. Most reports include line item costs plus construc¬ tion “escalators” to identify management fees and contin¬ gencies. Give projected cost ranges with projected fig¬ ures adjusted for inflation if the work will not begin within nine months. Photo Log: This section begins with a list of photos included in the report. Each photo should be captioned, but in the photo log, include a one-sentence description as well. After the photo log are the photos themselves. I prefer two 5×7 photos on a page to slightly magnify the details. Fulmer— 19 Biography: Conclude the report with a professional biography that describes consultants’ education and qual¬ ifications. Also include any credentials, professional des¬ ignations, etc. List any authored articles pertinent to the scope of the consultation. The inspection and reporting process can be chal¬ lenging and interesting. Each project has different requirements and may require altering some of the for¬ mat and processes discussed here. But the basic intent remains the same. As industry professionals, consultants must use their expertise in a fair, accurate, and objective manner to deliver the best quality product to clients. They should expect and receive nothing less. Fulmer— 20
