Adding a new plant feed to an existing substation

[rockman7892]

We currently own our substation which recieves power from the utility at 240kV and steps it down to 4.16kV via a 15/20/25 MVA Delta-Wye transformer. I have attached a 3-Line diagram of our existing substation, showing all of the equipment, and relaying associated with this substation. The bottom of the drawing shows where the cables terminate onto the substation bus in order to feed the incoming breaker and 5kV lineup on our existing plant. The existing plant has about 18MVA of connected load.

We are proposing building a small grinding plant here on the same site and are in the preliminar stages of looking into how we will power this plant. Right now I am investigating the option of powering this new plant off of our existing substation provided there is capacity. We are in the preliminary planning stages, and I wanted to see if it is even feasable to tap of our existing substation before we press forward with any hardcore engineering.

Based on the provided drawing and relay scheme I was wondering the following:

1) Can we tap off of the Transformer LV Bus at the proposed point indicated on the drawing to feed our new plant with a proposed connected load of 2MVA provided there is capacity (The capacity is a big if at this point)

2) If the proposed tap point is not feasable, is there another location on this substation arrangement that could be tapped to feed the new plant?

3) Based on the relay scheme, would any new circuit interrupting devices or relaying need to be added or re-arranged.

Again we are only in the preliminary planning stages and want to look at the feasability of this before we move on. I appreciate any comments, thoughts, ideas.

 

[davidbeach]

What codes govern? If NEC you may have more difficulty that if NESC. Electrically it is possible. Physically - don't know. Code wise - don't know.

 

[Zogzog]

Physically yes, very possible

 

[alehman]

Seems like you might want a switch and some overcurrent protection for the new tap.

 

[rockman7892]

We are governed by the NEC. Does the tap rule apply for MV installations as well?

 

[jghrist]

If I am reading the one-line correctly, there is no low side breaker. The transformer differential, the bus duct differential, and the feeder protection to the existing plant all trip the high side circuit switcher.

You would have to include CTs on the new line in the bus ducts differential. I assume that the new feeder will have relaying protection.

If you don't add low side circuit breakers, then you would lose both feeders for a fault on either feeder. Is this acceptable operationally?

 

[davidbeach]

I'm not sure I see bus differential, but it isn't clear what the lower 387 is being used for as only one restraint winding is shown.

How the NEC applies is kind of up to the AHJ and have familiar/comfortable they are with these types of installations. With full low side protection that trips the high side of the transformer, I'd say that the low side is properly protected and no low side breaker is required (not that there aren't operational reasons to want a low side breaker).

If there is bus differential, jghrist is absolutely correct that you need to extend it to include the new tap.

You probably want breakers on each feeder, may be there and just not shown, so that a fault on one doesn't take out the other.

 

[alehman]

The tap rule applies to medium voltage. The NEC doesn't allow high side breakers to protect low side feeders for delta-wye transfs. The main low side feeder you are proposing to tap is therefore a tap itself. You must have overcurrent protection ahead of the 2nd tap.

If this equipment is outdoors, the outside tap rule would apply which allows (single) taps of unlimited length. Technically you still can't tap a tap, but as David said, your AHJ will need to decide.

What does the 387 on the secondary do?

 

[davidbeach]

The code rule about having to have a breaker on the secondary is premised on breakers that have integral overcurrent sensing. If you have overcurrent sensing on the secondary and trip the primary you have provided everything that the is looking for from a performance perspective. The 387 on the transformer can do overcurrent for each winding.

Once up on a time I did a job, per NEC requirements, where I had a 4160[Δ]-12470Y/7200V transformer that was located over 200 conductor feet from both the high side and low side breakers and per article 705 I had to treat both sides as both primary and secondary. I put a diff relay on the transformer, all CTs within the transformer enclosure, and set it to trip both high and low breakers for any fault. No problem with the City.

 

[rockman7892]

The 387 on the lower part of the secondary is being used for both bus differential, and low-side overcurrent. The utility has told me that the one set of restraints can be used for both of these features in the relay. I mad a mistake on my drawing, and put the tapped new feeders before the second set of differential CT's which obviously wont work as others have mentioned, I will have to have the tap after this secdond set of Diff CT's.

If I understand Davidbeach correctly the tap rule applys to breakers on the secondary such as a molded case breaker or other type of breaker that only has trip sensing included inside the breaker itself. Because we are using CT's on the secondary to trip a switch on the primary, these CT's act or take the place of a secondary breaker since they are supplying the secondary overcurrent protection. Is this a correct basic understanding of the rule?

The main breaker for the existing plant is located on the main incoming section for the main swtichgear lineup located aprox 1000ft away. The new plant main switchgear lineup will also be aprox 100ft away and I'm assuming can also have an incoming breaker on the switchgear. This would mean that any fault on the buses of in the plants will trip the respective main breaker and only bring the plant offline. However like others mentioned a fault on either cable feeders to the two plants will trip the substation and therefore bring both plants offline. This is most likely unacceptable so I'm kicking around the idea of installing a main breaker for the new plant at the substation so if this plant has a fault on the feeders it will trip this breaker and keep the exisitng plant Online. I am working on a One-Line schematic as we speak.

I know it refers to some type of inspector but what exactly does AHJ stand for?

 

[davidbeach]

AHJ = Authority Having Jurisdiction. Also known as the inspector.

My take on the application of article 240 of the NEC, and other NEC overcurrent rules, to installations greater than 1000V is this: Have CTs connected to overcurrent relays at every point the NEC would require overcurrent protection. These relays need to trip one or more breakers such that the portion of the system that would be protected by a low voltage breaker becomes deenergized. In some cases you will deenergize more of the system than low voltage equivalent; such as tripping the high side of a transformer when the NEC would require tripping the low side.

 

[rockman7892]

The other factor that will determine weather or not I can tap another feeder off of this secondary bus is rated capacity of the secondary bus duct. The secondary Bus and secondary switch has a continuos rating of 3000A. The existing plant is still under construction, however it has a calculated connected load of 2893A. This does not allow the necessary capacity for the aprox 250-300A calculated for the new plant.

My question and decision now becomes how to add this new plant load. The way I see it, I have two choices, one being to upgrade the capacity of this bus duct which may include a signifigant cost. The other option I've heard is that most likely the actual load when the plant is up and running will be much less than the calculated connected load. I've heard and seen in the NEC that there is an exception that lets you monitor the actual load in a plant for some time and if it is seen that this actual load is less than the calculated load, then the actual load can be used as the connected load value thus allowing more loads to be added on the system. If this is the case, then how long must the actul load of the plant be monitored for before any type of decision can be made? How much load can be added after this actual rating can be found.

We have another plant that is almost identical to the existing plant being built which has been running for almost 10 years. This plant has the same 15/20/25MVA transformer. Can I look at the data from the past 10years on this plant to determine the actual load of this plant and add onto it? (Same bus capacity issue) Can I used the loading information on the plant that has been running to make a decision on adding load to the new substation transformer.

What are the rules that govern these types of decisions?

 

[rockman7892]

The other thing that I am realizing is that because the existing substation transformer is Resistance grounded (400A) the new plant will also be resistance grounded.

Are there any negatives to having a much smaller plant (2MVA) resistance grounded with a larger plant (18MVA) resistance grounded on the same transformer?

 

[alehman]

Secondary sensing tripping primary breaker - that may take some discussion with the AHJ. It should be acceptable, but they are not all terrific at "outside of the box" thinking.

NEC 220.87 permits demand load (times 125%) to be used for sizing. The rule says to use a 1-year period, and then has an exception that allows for 30 days if 1 year is not available, and if seasonal loads are accounted for. I don't think the NEC allows using data from one plant for another.

Obviously a ground fault in either plant will momentarily offset the voltages referenced to ground in both plants. Maybe not a concern, but your new plant will need to be designed as resistance grounded unless you derive a new ground with another transformer.

 

[alehman]

Not to be rude, but if this plant is being constructed it seems like you may want to hire a professional engineer who knows the codes to design this thing for you.

 

[rockman7892]

Alehman

We will indeed have a professional engineer design this plant if and when we decide to build it. As I mentioned, right now we are in the early brainstorming stages and I was asked to find out if this was feasable. I'm sure we will have to get some PE's involved before we go ahead and make any decisions however right now I'm just trying to gather all of the information I can to bring to the table. I am new to this so I'm hoping to learn alot by going through this research process so that one day I'll be a PE designing such systems.

 

[alehman]

Sorry, I misunderstood your statement that "We have another plant that is almost identical to the existing plant being built"