Contamination In SS 316L Pipe/ Jacket

Where does the makeup water come from....Is this a closed loop? Is there a water treatment system in the existing chiller loop? What is the contamination you speak of?

What is FYI?

A1. Why the contamination will be really significant if you bypass the PHE? Contamination is not significantly generated there, it will be present in water & it depends on source of water. MS pipe tend to corrode faster for chilled water service if outage are frequent, so little rust particles.

A2. Filters are definitely available but you should be certain, whether you really require it.

A3. Replacing the Mild Steel Pipe by SS will be definitely better for piping life point of view because MS pipe corrode faster. Copper is the tubing material for chiller (not the main pipe), it is not at all a issue here.

It depends on whether the Process Buffer Tank, and whatever is downstream of it, can cope with small doses of corrosion products from the mild steel piping.

• What on earth is the economic justification for removing the Plate Heat Exchanger?
• Given that its removal could lead to contamination of the Process Buffer Tank, why does the facility need to do this?
• Why was the Plate Heat Exchanger designed and installed in the first place, if there is no perceived need for it now?

HI,

Fisrt of all recommend against removing the phe. The reason the loops are seperated are to avoid scale formation in the chiller, which can be very difficult to remove. The phe is much easier to forstly monitor for loss of heat transfer efficiency, and also to clean.

You need to be very sure of your reasoning to remove the phe. Can you explain why you want to remove the phe?

Regarding the question of filters to remove contaminants. This depends on what the contaminents are, assuming you are talking dissolved metals (Fe++, Cu++, etc) then it gets difficult. Reverse Osmosis will remove them, but this is very expensive option for you. Suggest much cheaper to leave the phe in place.

It is hard to say how copper (or other contaminants) will cause problems ir not. Really need to understand the water chemistry involved, and exactly how the water is used. Also need detailed P&ID's of the system. Without this info, difficult to accurately give advice.

Hope this helps.

Anthony

You have not mentioned ANY use of DI or similar demineralized water in the plant loop, so I assume that there is none.

You are currently running a plant cooling loop at 2-3C deltaT, with relatively "warm" water. I assume (not stated) that the buffer tanks are due to variations in loads.

Your BEST solution is to replace the piping to and around the chiller with stainless- just to keep it all the same and eliminate the corrosion that ALWAYS happens with iron pipe AND eliminate BOTH buffer tanks.

Connect the chiller with a simple loop, with the pump feeding water into the chiller (so the chiller removes the pump's "waste heat"). Connect the piping from the plant to the chiller piping in a primary-secondary manner- Flow through the chiller is constant, with a flow rate matching the (peak flow x deltaT) loading of the plant loop relative to chiller capacity. Set the chiller's leaving water thermostat for the Plant's design temperature. Use VFD drives on the main plant feeds and two-way valves for ALL process loads of the plant loop.

To make this concept as simple as possible, let's assume that the chiller is sized for about 2100 kWt and that the plant PEAK load is also about 2100 kWt but can be as low as 1200 kWt. Based on the defined maximum 3C differential for the plant (14-11C) the plant's peak water flow load would be around 10,000 l/m, falling to as low as 5,700 l/m. The chiller pumping load would be 10,000 l/m constant.

The primary-secondary installation uses two connections to the chiller main (primary) loop, installed about 400-500 cm apart (as close as two TEES directly connected on the runs) with the return side "downstream" and closest to the pump feeding the chiller.

The pumps feeding the plant extract water from the primary loop (at the design low temperature) and feed it through the "secondary" loop to the users. Any "primary" water that is not extracted continues on toward the pump and back to the chiller. The water flow in the primary loop between the secondary extraction and the secondary return will ALWAYS be less that the primary flow- as low as zero (but not usually that low). In the worst possible case, where pump flows are not matched, it is possible that the primary flow between the two secondary lines could even be reversed, with a small amount of return water mixing with the cold supply water. That does not usually happen because the chiller is sized for the peak load of the plant which almost never occurs simultaneously in the secondary feeds.

That way, design cold water is always fed to all of the process users in the plant- as they need it to support their loads- and the warm return water is always either fed directly to the chiller or is mixed with the unused chiller supply water.

The advantage of THIS arrangement for the plant is that the chiller will be seeing warmer return water, and making warmer supply water, than it is now- leading to an increased capacity and increased efficiency- lower operating costs. The second benefit is that BOTH the chiller and the plant pumps will see significantly reduced operating pressures due to the elimination of the heat exchanger losses- you will likely want to use a VFD on the chiller pump (running at a fixed speed) also (it is cheaper than changing the pump). Finally, the VFD drives on the secondary feeds will also yield reduced electrical usage. An added value is that, with the 2-way valves and VFD, the secondary circuit will show nearly constant deltaT, not the ranges indicated in the sketches.

By the way, even IF the water in the plant feeds is DI or equivalent, there will be virtually no problems. I have used such a system feeding chilled DI water through stainless piping to copper electronics in a process furnace before with no contamination or copper loss issues- we eliminated iron plating that was causing an issue.