DHW System Recomm.B2CE, Inc_Consultant Report_6-13-2015.pdf

Building System Solutions Page 1 of 14

B2CE, Inc.

P.O. Box 17030 Golden, Colorado 80402

(303) 526-1800 (main)

(303) 526-1805 (fax)

Consulting Engineers

June 13, 2015

Mr. Cary Pilon, P.E.

Project Manager Colorado Department of Technical Services

VIA EMAIL

Mr. Al Cole Mr. Doug Park Colorado Department of Human Services Division of State and Veterans Nursing Homes McCandless State Veteran’s Nursing Home

VIA EMAIL

Re: McCandless State Veterans Nursing Home – Domestic Hot Water Investigation Part of State of Colorado Contract 74431-IHKA-FLNH1415-1 903 Moore Drive, Florence, Colorado B2CE, Inc. Job No.: 14010.02

Dear Al, Cary and Doug:

At your request, and per the above contract, I have prepared this report regarding my investigation into the domestic hot water generation and distribution system for the McCandless State Veterans Nursing Home.

I also provide my recommendations to correct the problems I found. I issued a preliminary report for review (dated April 25, 2015). Some recommendations in that report have been implemented and this report includes those changes (where know).

Scope of B2CE Work

I have been asked to diagnose the underlying cause of problems with the domestic hot water system for the building.

Mr. Al Cole, Cary Pilon and Doug Park

To date, I have performed the following activities:

I have performed several site observations, starting (informally) in July of 2014. I reviewed the observable as-constructed details of the existing system.

I conducted temperature and data logging of the system’s performance for a little over two weeks in late November, early December of 2014.

I have reviewed the installation and engineering information for the installed domestic hot water system components.

I have determined the factors that I believe are causing the problems.

I have described and sketched recommended modifications to the system to minimize the current problems.

Limitation

This report contains my professional opinion regarding the building systems I have reviewed to date. My opinions are based on my site observations, the documents I have reviewed so far and my knowledge and experience as a mechanical engineer. I reserve the right to modify my report and its opinions pending further review of additional information, should that information become available to me.

A. Building Overview and Background

The domestic hot water system has been revised several times since the building was first built. The current system was originally designed as part of Colorado State (Mechanical) Project Number IL9603 in 1996.

The system is a “dual temperature” system with separate hot water generators and storage devices. The current design intended to generate hot water at 180º for storage in two 175 gallon tanks – creating about 270 gallons of “usable” high-temperature hot water.

This system, which uses dedicated gas-fired “volume water heaters” (“boilers”).

This system replaced a similar dual temperature system (original with the building) that used a much larger storage tank with a large built-in electric heater (of unknown size). The current boilers sum to about 680 MBH heat output at altitude.

Hot water (“Domestic hot water” – DHW) is used at the kitchen and laundry. The laundry’s temperature requirements for the hot water have changed since the system started using an Ozone generator disinfection system. The Ozone system works best with cold water. My investigation did not include the suitability, ventilation or connections related to the laundry or its Ozone system.

Changes in the kitchen equipment are unknown, but it is assumed that required temperature is at least 160º (which is typical of commercial kitchens).

The DHW is mixed with cold water (DCW) at a dual-valve, “tempered mixing valve assembly” or “TMV-1”. The purpose of TMV-1 is to create relatively low-temperature “tempered hot water” (THW) – for use in public and residential parts of the building.

With a TMV assembly, high temperature DHW can be generated and used where necessary, in non-public parts of the building (like the laundry and kitchen). The higher temperature water also kills bacteria (like legionella) and reduces the risk of spreading bacteria through the DHW system.

When properly installed, maintained and operated, a TMV assembly also nearly eliminates the scald hazard that exists with higher temperature water in “public” and residential areas of the building. The original re-design anticipated supplying 120º THW to the building.

Each hot water system used two small pumps to draw water from the end of the DHW and THW distribution systems (via “hot water circulation” – HWC – piping).

HWC pipes route back to the mechanical room.

As hot water is used out at the fixtures, it “induces” cold water “make-up” to flow into the system. The re-circulated (HWC) water is intended to mix with domestic cold water (DCW) to provide the “make-up” water flow.

In most systems, make-up water is injected near the storage tanks such that pressure fluctuations in the system have the minimum possible effect on cold water distribution to the rest of the building.

Instead of being injected into the tanks, make-up water is currently injected into the “generation loop” on the pressure-side of its pump, between the boilers (“generators”) and the storage tanks. The injected water interferes with the hydraulics of the generating loop. There is no “hydraulic separation” between the make-up pressure changes and the generating loop pressure changes.

The HWC pump for the THW system currently injects re-circulated water into the cold water make-up line and has a branch connection directly to TMV-1. The HWC pump for the DHW system currently injects re-circulated water into the generating loop (between the two boilers).

The THW HWC pump is always drawing 120º water from the THW system and injecting it back into the make-up water, where it is intended to flow to the storage tank (during high use), or directly to the TMV (at no or low THW use).

The DHW HWC pump is intended to inject its (high temperature) water into the storage tanks (only) such that it does not raise the temperature of the THW.

Because of its injection point, this is not occurring at present. Instead, the high temperature HWC is injected into a point in the system where it must “fight” the pressure created by the generating loop’s circulating pump.

The current system has relatively new boilers with dedicated pumps. They are intended to divert a large portion of water out of the generation loop, heat it, and then “inject” it back into the loop. This is called an “injection loop”. Theoretically, if the tees for the boiler supply and return are close together, they don’t influence the “generating loop” or other boiler’s injection loop’s pressures. This is an example of “hydraulic separation”.

There is adequate hydraulic separation between the boilers and the generating loop. However, because of the orientation of the injection loops, it is hard to “lead/lag” the boilers and get maximum capacity out of the system.

B. Recent Changes to System

The original boilers were replaced about five years ago. The replacement boilers quickly failed. They were replaced about 2-1/2 years ago (with manufacturing dates of September 2012).

Other major components are “original”. They include the TMV assembly and the two storage tanks. The storage tanks have date-codes on their nameplates that indicate they were manufactured in 1993. They also have sacrificial anodes at the top of the tank (see the attached piping diagrams), which have never been serviced.

The system has been problematic since original installation and a number of piping modifications have been made in an effort to minimize problems. These modifications include:

Adding check and balancing valves in various locations;

Making various modifications to the HWC piping to its two termination points (TMV-1 and the make-up loop);

Relocation of the CW branch feed to TMV-1 (and adding additional check valves);

Installation of a drain-timer (to periodically remove “too hot” THW from the system);

Reducing the domestic hot water generating temperature to as low as 135º (from 180º design);

Reducing the tempered water temperature (from 120º design);

Installing an automatic shut-off valve at the discharge of TMV-1 (controlled by sensed THW supply); and

Installing a remote THW HWC circulation pump (called P-2 on my sketches) near the “far-end” (north) of the building – where the three branches of the original THW HWC circulation piping join together.

Since this pump was installed, the new Dementia Wing (constructed a few years ago) tied into the existing domestic water distribution piping. In the process, a new HWC piping line was connected downstream of the new HWC pump. It ties into (what has become) pressurized HWC (from P-2).

None of the fixes were completely successful, but the staff has learned to “manage” the system and is able to generally keep THW temperatures in check – except very late at night, when there is little usage.

A. Lack of Hot Water Capacity

The TMV effectively increases domestic hot water “capacity”.

A single (120º) temperature system would have to generate and store 100% of the highest demand for hot water.

The dual temperature system design intent is to reduce the demand by mixing cold water with DHW to make THW.

o If the storage tanks stored water at 180º, the system would only use about one half1 as much high-temperature DHW as it would if the water were generated and stored at 120º.2 In other words, the available storage capacity would go “twice as far”.

I did my data logging with the DHW control set to 145º, but with the actual control temperature at about 141º3. The data logging showed an inability to maintain DHW temperature at higher usage. The DHW temperature dropped to as low as 126º and to an average of 135º for as many as four hours at a time.

I believe the major reason for this temperature drop is that the available storage capacity is being used at too great a rate.

o The operating condition I logged (with 145º nominal DHW setpoint) effectively reduces available storage capacity even further.

o With 145º, the DHW demand to make THW is increased to 74% to 79% (depending on DCW temperature).

The difference in DHW demand to make THW (from the design intent to the current condition) effectively requires bigger boilers or more storage to keep up with peak tempered and hot water demand.

A higher “hot water” (DHW) temperature also reduces demand at the kitchen and laundry. This demand also results from mixing cold water with (high temperature) hot water – but at the fixture (out in the system). For instance, if the kitchen dishwasher is set to deliver 150º water, it will mix 80% DHW (at 180º) with (say 50º) cold water to make 150º water.

However, if the boilers are set to provide 150º (or lower) hot water, the dishwasher will use 100% DHW.

B. Domestic Hot Water Demand Calculation

I performed a cursory hot water demand calculation using recommended estimating methods in the International Plumbing Code (IPC) and other sources.

My calculations were based on counted THW fixtures on the recent remodel drawings, but I had to estimate the kitchen and laundry loads.

1 Actual use depends on cold water temperature which typically varies from about 60º in the summer to about 40º in the winter.

2 With 180º DHW and 120º THW, and assuming 60º (summertime) DCW temperature: This effectively reduces the tempered “hot water” (120º) demand to 50%. In the winter the deomand is “only” reduced to 57%

3 By Doug Park’s report, the DHW setpoint, which was once reduced to 135°, has been increased to 150º, now.

1. With 180º DHW Setpoint:

I found that the calculated maximum (180º temperature) DHW demand on the system is about:

a) 79 gpm (instantaneous maximum demand);

b) About 48 gallons per minute (gpm) for 12 minutes (semi-instantaneous demand); and

c) A maximum hourly demand on the order of 800 gallons per hour (gph).

The existing boilers can generate as much as 6304 gph. Therefore, if the system is operated at 180º:

The additional hourly demand of 170 gph (800 – 630 gph) can easily be met with the existing storage capacity.5 But, The continuous capacity of the boilers is 10.5 gpm – so the calculated draw-off from the tanks might be 36 gpm (48 gpm – 10.5 gpm) for as long as 12 minutes. This equates to 450 gallons of storage required (36.5 * 12). The actual storage capacity of the system is about 2606 gallons (usable).

Therefore, by calculation, the system may be too small, even when DHW is generated and stored at 180º.

However, calculated loads are generally very conservative. There is a good chance that the system is adequately sized if operated at 180º.

2. With 141º DHW Operating Temperature:

However, as explained above, the lower operating mode requires more storage. In other words, the system is likely to “run out of hot water” more often with a lower DHW operating temperature. At a DHW operating temperature of 141º (similar to what existed during the data logging), the storage demand is increased by about 30%.

With a 141º de facto DHW temperature, the maximum hourly demand is increased to 1100 gph and 62 gpm over 12 minutes.

If the boilers are able to deliver their full rated output, the storage tanks

4 Calculated at a 130º temperature rise.

5 Of up to 260 gallons of high-temperature water at one draw-off.

6 350 gallons of gross storage at 75% usable capacity.

would need to provide 470 gallons of usable storage. This is significantly less than the 260 gallons of usable storage available. Therefore, on an hourly basis the system becomes much more marginal, And, the 12 minute demand increases the calculated draw-off from the tanks to 51 gpm – or 600 gallons over the 12 minute demand period. The storage is certainly insufficient to meet the 12 minute demand.

Also, the current configuration of the generating loop controls is not likely to demand 100% fire rate from both boilers (for the entire duration of a 12 minute demand). This is because the boilers are currently controlled by the bottom-of-the-tank temperature and may not sense a demand until the storage is depleted. The system could be made more responsive by controlling the HWS temperature (back to the storage tanks).

My calculations predict that the system will lose control when operated at 141º operating temperature, and the data logging confirmed this. See section 0 below.

C. Lack of Control over THW Temperature

The main issue is a lack of control of the THW as it leaves the mechanical room.

It fluctuates far more than the tempered mixing valve assembly should allow.

THW climbs over-night when usage is low. THW drops during the day, when the demand is high.

Fixtures on the system may experience cold, appropriate or (occasionally) near-scalding temperatures.

Reducing the actual operating temperature to 141º7 (to avoid scalding the residents) effectively reduced the useable capacity of the system (on both the high temp and tempered systems). This makes it harder for the system to provide 120º water at high demand (when the DHW temperature gets out of control).

Inappropriate HWC (both DHW & THW) and DCW injection points; lack of hydraulic isolation; and improper balance of flow also contribute to poor THW temperature control.

Since my preliminary report, the TMV assembly has been serviced, and the DHW setpoint increased (currently it is at 150º). By report, system performance has been better.

However, there is evidence that the THW HWC flow is not properly balanced.

There appears to be a blockage in the branch connection to the storage water.8

7 or 135º

8 The Leonard representative visited the site on June 2, 2015. This is the condition per his report.

The system must be balanced such that 70 to 80% of the HWC flow is directed to the TMV assembly (with 20 to 30% of the flow to the make-up loop) – at high THW use. The generating loop and the high-temperature HWC must not be allowed to heat the THW HWC loop.

All of the other measures I recommend are necessary to further minimize the scalding risk to the residents.

Increasing the DHW setpoint to 160º (or higher) may be necessary to maintain DHW and THW delivery temperatures at high use.

D. Velocity Erosion

Many of the pipes exhibit external accretion of minerals, especially at fittings.

This is a sign of poor workmanship and/or excess water velocity in the pipes.

See the attached photos.

I found that the likely maximum velocity in the piping exceeds the recommendations of the Copper Development Association (attached).

The high-temperature DHW piping that feeds TMV-1 is especially prone to velocity erosion – especially when the system is operated at reduced DHW temperatures. The pre-manufactured inlet piping for the TMV-1 assembly is also too small, but is required to be so to fit the assembly inside the enclosure (and to match the inlet sizes of the mixing valves). The existing electric high-temperature shut-off valve also creates very high velocity (and the resulting pressure drop may further compromise operation of TMV-1).

I recommend that undersized piping be replaced with larger pipe, or with threaded brass piping, to reduce velocity erosion. I recommend that the electric shut-off valve be removed.

E. Maintenance of System

The staff has installed a number of system modifications in an attempt to address the poor temperature control in the THW system as discussed above (in section II above). The maintenance of the system is made more time consuming and difficult by the hardness of the water. The TMV assembly has to be re-built several times a year.

Piping has been replaced (and will continue to be replaced) due to velocity erosion.

I monitored system temperatures in several locations in the system from 4:00 pm on November 20 through 11:00 am December 6, 2014. I also monitored system pressure at the sink in Office 102 (Kathy’s Office) – a plumbing fixture far from the DHW system plant (and cold water entry).

Representative graphs are attached to the end of this report. They show:

A. The Generating Loop Temperature Varied Too Much

See Figure 1. The HWR (from the tanks dropped from an average of 141º to a low of 125º. Its maximum temperature was 145º. The temperature drops when large amounts of cold water make-up are mixed with the hot water (HWS) generated by the boilers. The data confirms frequent, but short loss of temperature control (as predicted by the load calculation.

At low use, the “generating loop” kept up easily with the demand (and maintained very close control). The HWS stayed close to the HWR temperature, generally over-shooting or under-shooting the setpoint by only +/- 2ºF.

However, at times of high demand (multiple times throughout the day, but specifically in the morning and at lunch time), the HWR/HWS had a hard time maintaining the tank setpoint temperature (aquastat set to 145, but de facto setting of 141º).

B. The DHW Varied Too Much

See Figure 2. The DHW temperature essentially mirrors the HWS. At low demand, the DHW was maintained within 1º (see performance at night).

However, at high demand: The generating loop temperatures drop, which causes the DHW temperature to drop. The maximum DHW temperature was 145º, averaged 141º and dropped as low as 126º (at high use). When averaged over 12 minutes, the typical loss of control dropped the average DHW temperature 11º (to 130º). Three other high-use loss-of-control events dropped the DHW temperature as low as 126º, 126º and 127º (average over 12 minutes).

C. The THW (and DCW Feed to the TMV) Varied Too Much

The high-end temperature control of the THW is relatively insensitive to drops in DHW temperature. The THW temperature generally stayed above 117º, even when the DHW dropped to 125º. TMV-1 performed well despite fluctuations in DHW supply temperature.

However high DCW temperatures had a profound effect on the THW temperature. See Figure 3. The DCW was measured at two locations. One (which is shown on most of the graphs) was measured near the TMV on the 2” line that feeds the TMV. This is the purple line on the graph. It “averaged” 100º ( ! ) It ranged from 55º to 129º. It was hottest at times of no THW demand. This is strong evidence of backflow through the valve (as discussed above). As discussed, TMV-1 was re-built after my preliminary report, and performance is reportedly improved.

However, it is important to note that: If the DCW temperature is too high, the THW will be too high.

For example, on November 21st from 11:00 to 5:00 on November 22nd, the average DCW temperature was 127º +2/-14º; the DHW was 142º +0/-2º and the THW was 131º +2/-8º. See Figure 5.

The DCW temperature was also measured at the entry in the room (through the chase on the south). The temperature trended down slightly during the two weeks of testing (as the average ground temperature dropped). Here the temperature varied from a minimum of 53º, with an average temperature of 56º and a maximum of 61º. This temperature is much closer than what comes “out of the ground” and was not affected by the DHW system.

“Something” is causing the DCW feed to the TMV to get hot. That something is the current connection between the cold water make-up and the generating loop;

and the high-temperature HWC and the generating loop. Injection of high-temperature HWC into the generating loop “over-pressurizes” the make-up with hot water. The make-up loop (including both HWC systems) must be hydraulically separated from the generating loop. This will occur if the existing expansion tank is properly located, and if the make-up connects to the storage tanks directly (as shown on the attached sketches).

D. The THW Varied Too Much

Thermostatically controlled mixing valves are tested to a temperature difference (between in-coming hot and cold water of 80º, and at the same pressure (+/-1º).

They can handle steady pressure imbalances of up to 20 psi and still maintain a discharge temperature +/-3º. As the in-coming temperatures get closer, and/or if the incoming pressures difference exceeds 20 psi and/or if the pressures fluctuate rapidly: the discharge temperature may get out of control.

I noted rapid pressure fluctuations at the sink in Kathy’s office (see Figure 6).

These pressure differences will be seen throughout the system, but are generally diminished at TMV-1.

I noted an average in-service pressure difference of 0 to 1 psig between the hot and cold water at low-use periods. I noted an “average” pressure difference of up to 16 psi (40 to 56 psi, respectively) at high-use. The pressure difference is almost entirely caused by flow through the TMV. A 16 psig pressure drop (in the TMV) equates to an instantaneous flow of approximately 50 to 60 gpm. Most of this pressure drop is unavoidable, but could be reduced with a new TMV assembly.

The THW varied from a minimum of 114º to a maximum of 133º, with an average temperature of 124º. One can see the effect of “cold” and DHW temperatures on THW in Figure 5

1. THW Control May Be “OK” At Present

As long as the DCW feed temperature remained at a reasonably low temperature, the THW was in control. When the DCW temperature was “in range” (i.e., less than about 75º), the THW was maintained between 115º and 117º. This is very acceptable control and probably represents the current operating condition – despite the rest of the plant being far from optimized.

For example, during “after lunch clean-up” from 12:30 to 1:30 pm (when the laundry was presumably operating, too) on November 22nd: The average DCW temperature was 65º +36/-8º; the DHW was 135º +3/-2º and the THW was 118º +8/-4º.

E. DHW System Pressure Drop Exceeded Cold Water Pressure Drop

See Figure 6. Pressure fluctuations at the subject fixture were significant and rapid. The DCW cold water pressure was almost always higher than the hot water – and the THW pressure had the greatest variability.

The average DCW pressure was 58 +26/-18 psi. The average THW pressure was 48 +39/-29 psi.

For the “busy” time from 12:30 to 1:30 pm on November 22nd: The average DCW pressure was 56 +9/-7 psi. The average THW pressure was 42 +18/-13 psi.

However the simultaneous difference averaged 9 psig and the moving average over 32 seconds did not exceed 26 psi.

These pressure differentials are generally acceptable for proper THW response (except for rapid pressure fluctuations). However, they confirm that the TMV and tempered hot water pipe sizing are somewhat undersized.

See the attached sketches. They summarize my recommended changes. All of the changes should be implemented at the same time. There will be minimal improvement with “piecemeal” modifications to the plant

In general the recommendations:

Provide hydraulic separation between the generating loop and the make-up system.

This should minimize hot water backflow into the cold water line.

Provide the correct piping arrangement for the DHW & THW HWC piping.

At low demand rates, THW HWC will flow into the by-pass port on the valve.

Together with hydraulic separation, this will prevent over-heating of the THW.

Except for TMV-1, replace components that have likely reached the end of their service lives.

A. Inspection of Anode Rods and Possible Replacement of Tanks

The anode rods for the existing storage tanks are likely inaccessible. This means that they have probably never been serviced. Once the sacrificial anodes are “used up”, corrosion attacks any micro-cracks in the glass liner. As a result, corrosion of the steel tanks tends to concentrate at the cracks, and corrosion is accelerated.

I recommend replacement tanks (of at least the same capacity) – but with access to the anode rods. The new tanks also require an additional full-size connection fitting (compared to the existing tanks). Replacement tanks are shown on the attached sketches.

B. Replacement of TMV Assembly

The existing assembly is very high maintenance – specifically due to high use and water with high mineral content. The State should consider replacing the assembly with a single valve system that is less susceptible to fouling due to accretion build-up.

Two potential replacement options are valves manufactured by:

Heat Timer (for example, see http://cdn.heat-timer.com/wp-content/uploads/ETV-Plus-Lit-056192-A-Color.pdf), and

Lawler (for example, see http://www.temperedwater.com/wp-content/uploads/M805-Ins-Sheet.pdf).

C. Operation of DHW Generating Loop

The system is intended to operate at 180º DHW. There are some advantages to prolonged equipment life at a lower DHW temperature and maintenance staff experience seems to show that 180º may not be necessary.

Increasing the DHW setpoint temperature to (at least) 160º for best capacity control is recommended. Making this control change will increase storage capacity and allow full boiler output at high-load conditions.

However, this change should not be made until (at least) both the THW and high-temperature DHW HWC changes have been made. The higher operating temperature will not increase scald risk if these (and preferably all of the other recommended) changes are made.

The system in the boiler room should be fully insulated.

END OF REPORT

Respectfully submitted, B2CE, Inc.

M. Robert D. Barrett, P.E.

Principal

Attachments:

Sketches of Recommended Demo and Modification to the System (3 pages) Selected Data Logging Graphs (6 pages) Selected Photos (6 pages) Copper Development Associations recommendations for Velocity in Copper Pipe (2 pages)

X:\B2CE\Jobs\Florence CO Veteran's Nursing Home Invest.14010.00\14010.01.DHW Investigation\Report\Letter Report Florence VA Mech.DHW System Recomm.B2CE, Inc.20Apr2015.docx / 6/13/2015 8:26:00 PM / 6/13/2015 9:09:00 PM