Cleaning the Bookstacks After a Major Fumigation

Joyce Watson


Vacuuming half a million books wasn't exactly what we had in mind when we discovered serious insect infestations in the Hamilton Library stacks. When infestations had occurred in the past, we'd boxed up a couple of ranges of buggy books, sent the boxes out to be fumigated, then unpacked the books, stamped them with the date and "fumigated", and returned them to the shelves. This time, the infestation was obviously much worse. Staff had begun noting frass (insect droppings) in many different places in the bookstacks, and on several floors. Sometimes there was a thick layer of frass where it had poured out of the books onto the shelf.

In response to this growing problem, the library asked the Pacific Regional Conservation Center to conduct a survey documenting the environmental conditions in Hamilton Library that support insect infestations. The survey concluded that Hamilton's deteriorating air conditioning system was responsible for high humidity (often in the 70's and 80's), hot spots with temperatures in the high 70's and 80's, poor air circulation, and inadequate dust filtration--all factors that favor high rates of beetle propagation and survival. Fortunately, the University Facilities Management Office had been aware of the serious air conditioning problems at Hamilton, and arrangements were in the works for a partial air conditioning renovation. A final renovation, which promises even better climate controls, cannot occur until after the new addition to the library is built.The PRCC report1 made clear that our practice of returning fumigated books to the shelves without cleaning them was unsatisfactory. The insect frass, which we had been leaving in the books and on the shelves contains a chemical pheromone which attracts flying adult insects, thus encouraging a repetition of the cycle. Besides that, dust itself was part of the problem. Dust holds moisture and food particles, thus creating an ideal setting for book beetles. Even if we'd been able to wave a magic wand over our poor air conditioning system and get instantly improved dust filtration, we were stuck with the layers of dust that were already on the books and the shelves.

Hamilton Library has excellent custodians, but the crew is simply not large enough. One custodian per floor (from 25,000 to 50,000 square feet!) can hardly do more than wash the restrooms and drinking fountains, empty the wastepaper baskets, sweep the floors, and vacuum the carpet. Vacuuming the bookstacks has never been the responsibility of the custodians. Alas, as in most libraries, it is nobody's job.

At the very least, two things had to be done. First, we had to kill the book beetles, their larvae and eggs. And then we had to clean the treated books and the shelves. Ideally we would have cleaned all the bookstacks; realistically we could hardly imagine cleaning the 2500+ bookstack sections that held infested books--almost ten miles of shelves!

With the development of new preservation techniques, the method of choice in dealing with insect-infested books is to freeze them, wrapping the books in heavy plastic bags, sealing the bags, and bringing them down to freezing temperatures to kill all forms of the insect. The treatment works, it does no damage to the books, and it is non-toxic to humans. However, the sheer number of books needing treatment ruled out this procedure. Instead, we decided to fumigate. Hamilton Library was closed for four days over the Labor Day weekend, while fourteen areas on five floors were tented and fumigated. Infested books that were not in these areas were boxed and placed in the aisles of the areas that were tented.

The prospect of vacuuming half a million books, and the shelving on which they stood, was daunting. At first, handling so big a job in-house seemed impossible, so we invited several representatives from Honolulu's janitorial service companies to give us informal estimates. After they each received the same description of the job to be done, they submitted their estimates, which varied from $7,500 to $47,676! Presumably, reality lay somewhere in between those two extremes, but we didn't have a clue where, and we feared going out to bid and being saddled with a company that had seriously underbid the job. We were also worried that there would be hundreds of books, maybe even thousands, needing repair, and having them handled by an outside maintenance crew didn't seem like a good idea.

A literature search turned up one short summary of stack-cleaning procedures.2 Another article discussed a large bookstack cleaning project at the New York Public Library3. It was nice to be in such august company, but we needed a lot of practical, how-to-do-it advice. Finally, we decided to tackle the job with our own staff and custodians. That way we could feel our way along, dealing with problems as they arose.

A project team was chosen, composed of the the Head of the Circulation Department and three of her stacks maintenance staff, the Building Manager, and the Building Planning Coordinator. Ordinarily, in a project of this scope, a library's Preservation Officer would be involved. However, Hamilton's new Preservation Officer had just begun in her new position, and was busy setting up the new Preservation Department.

In all, seventy-five people worked on the cleanup--twenty UHM custodians, 41 student staff members, and 14 civil service staff. Several librarians also volunteered their services.

At first, the custodians worked both Saturdays and Sundays, but after several weeks, exhaustion set in, and they cut back to Saturdays only. Several of the crew worked in Hamilton Library during their regular shift. Having our own custodians on the team was a real advantage--their good spirits were exceptional, and very practically, they knew their way around the building, knew where the electrical closets were when a circuit tripped, and had building keys. The custodians always worked 8-hour days during the weekend cleanup, as Facilities Management ruled out their working overtime during the week. Project team members set up for the custodians on each Friday night, getting the vacuum cleaners, extension cords, and other supplies and equipment ready. The custodians were paid from library funds, at time-and-a-half.

Student staff members were invited to work additional hours, at a flat rate of $7 per hour. Because the cleanup began early in the semester, many students were able to put in extra hours. This was a factor in choosing the Labor Day date for the fumigation. We received special permission from the Student Employment Office on campus for students to work up to 30 hours per week as long as they had a minimum 2.0 GPA and were in good academic standing. We made it clear, before they signed up, that they had to be willing to work in dusty, dirty conditions, that they should expect to handle infested books, vacuum insect frass from the books and shelves, and that they might have to do some lifting and moving of boxes of books. Most of the student staff who elected to put in extra hours worked in the Circulation Department, and were familiar with the infested areas. While working on the vacuuming team, they were supervised by the Circulation stacks staff.

Civil Service library staff were also invited to work up to eight hours per week in addition to their normal working hours. Their pay rate for this job, with the same duties as described for the custodians and student shelving staff, was time-and-a-half.

Before beginning to vacuum, new workers went through an informal orientation, sometimes individually, usually in small groups. A checklist of information they received follows:

• How the bookstacks were mapped, showing exactly which sections had infested books. We demonstrated the procedure--tipping back the books shelf by shelf, to see where there was frass--and showed them infested areas in the bookstacks so they would know what to look for when they began vacuuming.

• A look at the book beetles themselves (dead, in a cellophane wrapper), and live larvae in a book. We gave a brief explanation of the cycle of the beetle: the flying female insect lays eggs, often in the spine of a book, the eggs hatch into larvae, the larvae eat their way through the book, then turn into adult flying insects. The cycle begins anew--the beetles mate, then the female beetles moves into the spine of a book to lay eggs.

• Importance of clean-up. Aesthetic considerations aside, frass attracts beetles. Also, we need to be able to find new infestations by spotting frass on the shelves. Frass left over from the fumigation and cleanup would confuse the issue.

• Samples of books that need repair were shown to the workers. They were told to remove books like these from the shelves, put a rubberband around them, and place them on a nearby booktruck marked "Books --> Repair".

• Workers were encouraged to wear dust masks and gloves while vacuuming.

• The vacuuming procedure was explained and demonstrated.


The project team wrote vacuuming procedures, which were amended as we went along. Even with the written procedures, we found it helpful to cruise around the stacks when people were vacuuming. We picked up hints from them to pass on to other workers, and were able to set them straight when they were not following directions.

Enlarged floor plans showing ranges that were to be vacuumed were posted on range ends. The floor plans were created using

Macintosh MacDraft software. Having workers initial boxes representing each single-face section they vacuumed enabled us to keep track of what had been done, as well as who had done it.

Signs were placed in the lobby informing patrons that the post-fumigation cleanup would be going on for several weeks in the bookstacks. Small signs were placed in the stacks where people were vacuuming, "SORRY ABOUT THE NOISE. WE ARE VACUUMING THE BOOKSTACKS. PLEASE BEAR WITH US."

"Confetti books" was the name we gave to volumes that were so badly eaten by book beetles that they were literally in pieces. If one were to open up a confetti book and tip it upside down, hundreds--even thousands--of small pieces of the pages would fall out. Obviously it was not possible to vacuum these! Workers were instructed to put a rubberband around these books and place them on a booktruck marked (quite unrealistically, in these cases) "Books -->repair". Books with loose pages or spines that had come unglued, or otherwise in need of repair, were also placed on the booktruck destined for the Preservation Department.

"Thunking" the books on the booktruck shelf did get a lot of frass out. Workers were told to be sure to vacuum inside the front and back covers where there was frass still in the book. In order to avoid tearing the pages, or even sucking them up into the vacuum hose, the workers decreased the suction on their vacuum cleaner by opening the ring on the hose. They also placed one hand, with fingers splayed out, over the pages being vacuumed.

Where the infestations had been the worst, we checked back afterwards and found a lot of frass still on the shelves, which had obviously fallen out of the books since they had been vacuumed and returned to the shelves. We were mindful of the fact that frass attracts insects, but there was also another important consideration. After having so serious an infestation, we were determined to monitor the stacks more closely in the future, watching for new infestations. The clue to a new infestation is to find frass on the shelf. But if we left frass after the fumigation, how could a monitor, coming by next month or next year, tell whether the frass was old, or whether it was a sign of new infestation? Sometimes we can tell whether frass is old or new. New frass tends to collect in neat piles around the bottom of the book spine, and is a bit moist. Old frass often appears finer and dryer, and is scattered around the shelf. But one cannot always be sure.

Some of the books were so filled with frass that we just couldn't get it all out, even on the second go-round. We were concerned about "thunking" the books so much that they would need repair. Our Building Manager came up with an effective and amusing solution. Reasoning that the "frassy" books needed some major-duty jiggling, he brought in a hand vibrator (a "personal massager") and found it worked beautifully. One opens the book a bit to loosen the binding around the spine where most frass is lodged, then runs the vibrator slowly up and down the outside of the spine. The frass pours out the bottom. Besides being effective, vibrating a book is easier on the book than banging it many times on a flat surface.

In areas of the building where our poor air conditioning system had resulted in hot spots and high relative humidity, some of the bookstacks had mold on the books and on the shelves themselves. Bound journals were affected the worst, but some monographs, too, were moldy. An earlier survey at Hamilton had found that many staff working in the bookstacks were affected by the mold, with complaints ranging from itchy eyes and skin, sinus headaches, and runny nose, to hay fever and sneezing. Workers who were bothered by the mold were not assigned to moldy areas. For those who did vacuum moldy books, we purchased several special items, including wet-dry vacuum cleaners, special masks, and cover-up shirts. Workers were also strongly urged to wear disposable vinyl gloves.

The decision to do the job in-house turned out to be a good one. Not knowing what to expect, what the pitfalls might be, we were in no position to hire an outside company to do the job. We needed to work out the details as we went along, trying different techniques, and learning the best methods. This process continued all through the project. When the inevitable problems came up, the project team met and figured out what to do. Once the cleanup was working smoothly, there was no reason to change the game plan.

Timing was critical. Once the bookstacks had all been examined and we knew exactly where the infestations were, we rushed to have the fumigation done over the Labor Day weekend, before the semester was well underway, so that student staff would be able to put in extra hours of work. Ideally we would have fumigated just before the semester began, thereby inconveniencing fewer patrons, and fumigating when the largest percentage of books were on the shelves, but once the decision was made to fumigate, the preparation of specifications for the fumigation, the bid process, etc., required a certain amount of time. The library was closed over the three-day weekend, and also for one additional day. Staff members were given administrative leave for the extra day that the library was closed.

We needed more time to select and order supplies and equipment, and they didn't always arrive on time. For example, the workers used awkward six-foot ladders for weeks before the rolling kickladders finally arrived. Although a few students did work right up to the end of the project, as we got further into the semester, student staff hours slowed down, and we relied more and more on the custodial staff to do the bulk of the work.

Working with the University of Hawaii Facilities Management Office staff was a pleasure. They supported the project wholeheartedly, and helped in every way they could. Besides making it possible for the custodians to sign up for overtime work, they sent ladders for us to use, and even purchased Eureka vacuum cleaners to make sure there would be enough for the custodians.

Everyone appreciated the opportunity to earn better pay for their hours. In the process, we all became more alert to problems in the stacks.

The team spirit that developed during the clean-up project among all the workers--custodians,student staff, civil service staff, and librarians--was a real bonus.


The project team met weekly. There was always plenty of business at hand, but the meetings were also a good morale booster. We were in constant communication with each other during the week via electronic mail, with copies of our memos going to all the project team members.

A humorous T-shirt (a take-off on Arnold Schwarzenegger's "Terminator") was designed by one of our student staff members, and given free to each person who vacuumed over fifteen hours. T-shirts were also given to librarians who volunteered at least four hours of free vacuuming time. We selected only one color purposely, so the shirts would be more easily identifiable. Project team members wore their shirts frequently, as did many of the workers.

A public display was put up in the lobby of the library. Books with frass, confetti books, dead book beetles (dead, in cellophane wrap), even the vacuum cleaners and vibrators, were included in the display. The centerpiece was a "temperature chart" showing the vacuuming of 500,000 books as the goal; it was updated regularly. Lots of photographs were taken of the whole procedure, showing the fumigation, the vacuuming that followed, and including pictures of the workers on the job. The display was very popular, and got a lot of attention from patrons, as well as library staff and custodians. In addition to the display, staff received periodic updates, as well as appeals for

For the Saturdays and Sundays that the custodians worked, we furnished donuts to get the day off to a sweet start. Toward the end of the project, we threw a luncheon party for all the workers, and many custodians and library staff--students, civil service workers, and librarians--joined us.


Check the electrical outlets before the workers arrive. Are they live? Plug all the vacuum cleaners in where you expect to use them, and turn them all on to see if the circuits will support them. In our experience at Hamilton, three vacuums on one circuit frequently caused it to trip. You may need to consult electrical floor plans to see which outlets share a common circuit.

If you will be using several vacuum cleaners, mark the circuit numbers on each outlet; then if the vacuum cleaner stops, you can check the number at the outlet and go directly to the circuit that has been tripped. Sometimes in order to use several different circuits, we had to run cords from areas across the building, thus the need for 100-footer extension cords.

Set-up time. It takes longer than you think, especially at first. We spent six person-hours the first Friday night getting ready for a crew of fourteen! (Never start with a crew of fourteen--five or six workers is plenty when you're just beginning.) At least one project team member must be available while the crew is working--there's no predicting what may go wrong.

Forget using a plank! We thought long painters' planks would be wonderfully efficient for workers cleaning the top couple of shelves--that way just two or three people could do all the top shelves, then the rest of the workers wouldn't need to be going up and down ladders constantly. What a mistake! The planks wouldn't fit in the service elevator, and though we did manage to get one 16-footer up to the second floor via the stairway, there was no room to maneuver once there, and we couldn't get it into the stacks.

White poster board, taped to the top shelf of the booktrucks, works wonders when you're getting frass out of books. You can see the frass better, to know when you're done "thunking" or vibrating the book.

Disposable vacuum cleaner bags should be changed regularly. Don't let them get even half full.


The cleanup project went surprisingly well. It took four months and seventy-five workers to vacuum half a million books and ten miles of shelves. The cost of the workers, not counting the staff members who served on the project team, was $31,366--about six cents a book. Supplies and equipment added another $6,089.

Having become aware of the extent of the infestation in the bookstacks, and determined not to let it happen again, the library administration has funded a full-time position in the Preservation Unit. This new staff member has responsibility for monitoring the bookstacks to be sure that future infestations are caught early and treated.


Honolulu businesses listed below are ones where we found the supplies and equipment we needed, at lower prices than elsewhere. They have knowledgeable sales people with whom you can discuss your needs. Costs are variable, depending upon quantities purchased.

Booktrucks. One booktruck per worker. We used both large and small booktrucks. The large booktrucks have the advantage of a top shelf long enough--38"--to hold an entire shelf of books. The smaller trucks--31"--are more maneuverable. We avoided purchasing any additional booktrucks by conducting Friday night raids in staff workrooms.

Eureka "Mighty Mite" vacuum cleaners with disposable bag. This is a small canister vacuum, lightweight and very durable. They are available in 1, 2, 3, and 4 horsepower models. We bought the 2 hp models for under $100. The custodians brought their own 4 hp Eurekas from Facilities Management . There is little difference in suction between the models. The 4 hp model is for heavy-duty work, and it's less likely to heat up with long use even when the bag gets full. On our 2 hp models, we changed the bags well before they were half full, and never had any trouble with the motors overheating, despite long hours of use. An adjustable strap is a necessity. We used it to hang each vacuum over the end of a booktruck. (We can attest to the Eureka's durabiity because, before we got the straps, vacuum cleaners fell off the shelf of the booktrucks onto the floor several times. Even so, all six Eurekas made it through the project fully functional, with only one broken wheel.) Whatever vacuum cleaner you buy, make sure it has a long hose. Remember that standard stacks are 7'6" high, and if there are canopy tops that need to be dusted, you've got to reach even higher. The one attachment you'll use the most is the round dusting brush. The vacuums all seem to come with the synthetic plastic type, and the plastic bristles wear down quickly, though we did discover that with a round brush you can swivel the brush around in its setting so that an unused part of the brush can be used. Horsehair dusting brush replacements (round and square) cost just a few dollars apiece, and are well worth the money. They last longer and do a better job. We bought extras at the Eureka Company, 1300 Hart Street, Honolulu, telephone 847-2683.

There's also an "easy-reach attachment" on the market, a long skinny brush that is specially designed for (among other things) vacuuming the tops of books when they are sitting on the shelves. It's no better than the round dusting brush for vacuuming books when they are off the shelf, but if you're doing a general dusting of the books in place on the shelves, it's excellent. The Vacuum Cleaner Center (see above) has a supply. They cost about $15.

Rainbow wet-dry vacuum cleaners. These were specially purchased for vacuuming moldy books. With a disposable paper filter such as the Eureka has, a lot of air is pushed through the filter and out of the machine, thereby broadcasting mold spores all around the vacuuming area. By contrast, the Rainbow uses water for a filter. Mold spores are trapped in the water, and dumped down the toilet when the job is finished. As a result there are many fewer mold spores floating around the area being vacuumed. The Rainbows sit in a wheeled platform and roll along the aisle easily. Get an extra long hose that will reach the top canopy--ours were fourteen feet long. The Rainbows are expensive. New models cost over $1200. We paid less than half that for reconditioned models. We bought four, without the power nozzle for cleaning carpets, and were able to get a several-year parts warranty.

Sources for vacuum cleaners:

Discount Vacuum & Sewing Center

404 Ward Avenue

Honolulu, HI 96814

telephone 524-5450

Owner: Randy Wischnesky. Parts repairer par excellence--Phillip Felix

Vacuum Cleaner Center

1120 S. King Street

Honolulu, HI 96814

telephone 536-5147

Owner: Norman Schneider.

Extension cords. We used 50-feet and 100-feet, 12 gauge/16 amp extension cords. Fortunately they were bright orange and could be easily seen by patrons. There was no avoiding running the cords for long distances alongside stack ranges and even crossing main aisles. We attached theft-detection strips along one end, and wrapped the strips with masking tape. Write "50-feet" or "100-feet" at both ends. This makes for faster identification. To avoid having a big box of jumbled up cords (which takes forever to unscramble, especially if there are people waiting for them in order to get to work!), use plastic cord wrappers. They're shaped like an H with a double bar in the middle, and have hooks where you attach each cord end. Power bars were needed where two or more vacuums were in the same area and a long extension cord ran from one electrical outlet. We also used them when each worker was equipped with a vibrator as well as a vacuum cleaner. City Mill or any large hardware store has all of these items.

Vibrators/personal massagers. The vibrators should have both a high and a low setting. On more fragile books, we used only the low setting. Slip the vibrator into a sock to keep frass from getting into the motor. (Available at Longs or any of the large discount stores.) The library's fiscal officer gets the job of explaining to central purchasing why the library needs vibrators.

Kickladders. (Cramer 3-step ladders). A tall worker can manage with an ordinary library kikstool, but the stools aren't high enough for an average-height worker. You really need to see what you're doing. We purchased 3-step kickladders that work on the same principle as library kikstools, i.e. they roll until one puts weight on them, and they were excellent. Specify a double top step, which gives the workers a wide footing. The two local companies with the best prices were Office Systems Hawaii and Office Concepts Hawaii.

Dust cloths. Scraps of absorbent material will do, as long as the fabric isn't shiny--toweling, t-shirts, etc. We sprayed them with Endust to make them more effective. Sometimes the dirt stuck to the shelves right through the vacuuming. A good rub with the cloth was enough to get the shelves clean. We did not dust the books with the cloths.

Dust masks. 3M Comfort Mask #8500. Can be purchased at Gaspro. For vacuuming moldy areas, the 3M Dust and Mist Respirator #8710 model is recommended.

Plastic gloves. ("Tru-touch" gloves). Gaspro, again. Most workers did not choose to wear plastic gloves when they vacuumed frassy books, and we made it optional. It was really an aesthetic choice, except for those few who were bothered by the dust. However, when they were vacuuming in moldy areas, we strongly encouraged them to wear gloves. Even if handling moldy books did not bother their skin, wearing gloves reminded them not to put their hands to their eyes.

White long-sleeved men's shirts were worn by workers vacuuming in moldy areas. We purchased used ones at garage sales. They were worn for only one vacuuming session, then laundered before being worn again.

Joyce Watson, is at the University of Hawaii at Manoa in Honolulu where she is a Librarian, Building and Planning Coordinator. Email: joycew@uhunix.bitnet


1. Pacific Regional Conservation Center 1990, 19.

2. Swartzell, 88-89.

3. AB Library Note 1987, 1378-1379.


AB Library Notes. "Spring Cleaning at N.Y.P.L. Starts in the Stacks." AB Bookman's Weekly, March 30, 1987, p. 1378-1379.

Pacific Regional Conservation Center, 1990. Unpublished report. University of Hawaii, Hamilton Library, General Conservation Survey Report.

Swartzell, Ann. Preservation Column, RTSD Newsletter, 1985, vol. 10, number 7, p. 88-90.